• Colossal Biosciences is aiming to revive extinct species like the woolly mammoth.
• The startup has raised over $400 million to fund its de-extinction and conservation missions.
• Its founder and CEO, Ben Lamm, told BI about the potential revenue streams the startup is exploring.

How can a startup provide a return on investment from bringing the woolly mammoth back from the dead?

Ben Lamm, the CEO and cofounder of Colossal Biosciences, the startup developing "de-extinction technology," says the real returns are in the discoveries you make along the way.

The startup has raised over $400 million from investors for its headline missions of reviving extinct species and conservationism.

But in the same way that the space race gave us ancillary technological innovations like GPS and artificial limbs, Lamm says that Colossal Biosciences' mammoth moonshot could spur monetizable advances in biotech.

"There's so much that goes into humans landing on the moon," he told Business Insider. "Well, the same thing goes for de-extinction — you have to build computational biology, cellular engineering, stem cell reprogramming, genetic engineering, monoclonal screening, and embryology."

Some of those technologies could be "so powerful," said Lamm, that the startup is exploring spinning out technologies in areas like longevity, human healthcare, and plastic degradation.

But first, the startup must find a way to revive an animal that went extinct 4,000 years ago.

A mammoth task

So, how does one bring back an extinct animal?

Lamm told BI that the Texas-based startup is taking the opposite approach to Jurassic Park, in which fictional scientists use frog DNA to fill in the gaps in dinosaur DNA that have degraded over time.

Instead, Colossal Biosciences starts with the closest living relative to the woolly mammoth, the Asian elephant.

The startup then uses AI to analyze samples of woolly mammoth DNA, including some well-preserved samples from Alaska and Siberia, ranging from "about 3,500 years old to about 700,000 years old," Lamm said.

That helps Colossal "understand what makes a mammoth a mammoth" before the startup engineers those genes into an Asian elephant cell, Lamm said. In other words, it would be like adding dinosaur genes to frogs, rather than using frogs to fill in dinosaur DNA gaps.

"It's like reverse Jurassic Park," Lamm said.

Colossal is aiming to bring back a woolly mammoth calf, born to a surrogate elephant mother, by late 2028.

On the path to reviving the mammoth, the startup this week announced a new breakthrough in multiplex genome engineering: the birth of a "colossal woolly mouse." The process involved engineering mice so they exhibited mammoth-like qualities, such as thicker fur, that would enable them to adapt to cold environments, Colossal Biosciences said.

"It is an important step toward validating our approach to resurrecting traits that have been lost to extinction and that our goal is to restore," Beth Shapiro, the chief science officer at Colossal Biosciences, said in a Tuesday press release.

Startups find a way — to make money

Beyond engineering ancient DNA, Lamm said that the company's ambition — and thesis for its investors — had always been to spin out research and technology cases with broader applications, like human healthcare.

Launched in 2021, Colossal Biosciences has attracted a slew of high-profile investors, including celebrities such as Chris Hemsworth and Paris Hilton, as well as firms including At One Ventures and Draper Associates.

Its most recent $200 million round, led by TWG Global, brings the startup's total funding to $437 million — at a valuation of $10.2 billion.

Colossal's first spinout company was Form Bio, which raised $30 million to help scientists manage large datasets. Its second spinout company, Breaking, focuses on synthetic biology to tackle the issue of plastic degradation — and raised a fresh $10.5 million last year.

"It's fun for our investors, who get a ride along with all this other stuff," said Lamm.

'It's kind of terrifying'

Alongside its de-extinction technology, Colossal is creating tools meant to help conservation efforts for species that still roam the planet.

Depending on the severity of the climate crisis, the world could lose up to 27% of vertebrae species by 2100, a European Comission forecast from 2022 said.

"We need new tools in the fight against biodiversity loss," Lamm told BI. "It's kind of terrifying."

The company says it provides conservation technology, which it calls "de-extinction toolkits," to its 48 partners, which include the charities Save the Elephants and Re:wild. These include breeding programs that help preserve endangered species, especially those threatened by zoonotic diseases, its website states.

One potential revenue stream is offering "nature credits." Similar to carbon credits, the idea is that companies would buy into nature-positive commitments, which Lamm says could become "long-term annuity streams."

Although speculative — Lamm didn't share a timeline — it could offer Colossal a huge opportunity in the wider carbon credits market.

"While we believe there's a massive ecological impact in rewilding, just to be capitalistic for a second, we think that we'll make billions of dollars in annuities off of having animals healthy back in the wild now," he added.

Another potential income stream for the startup is ecotourism. Colossal says it's in discussions with governments over potential "nine-figure" contracts. However, reviving one mammoth is the startup's priority before it can entertain the idea of a Jurassic Park-style venture.

Trillion-dollar ambitions

Colossal has plans to open its flagship lab this year — and a slew of milestones it aims to hit.

Armed with its fresh funding, the startup plans to examine other avian and mammalian species it hopes to revive. Colossal already plans to bring back the dodo.

The startup is also developing an artificial womb to gestate mammoths and other species, such as northern white rhinos and polar bears. By 2026, the company's artificial womb team hopes to birth a mammal ex-utero.

Lamm is bullish about the upstart's growth trajectory.

"We believe this is a trillion-dollar company," he said. "And we believe this company will potentially have more impact than many other companies have had on the planet."

• San Francisco-based Breakout has come out of stealth with $3.25 million in seed funding.
• The startup has developed an AI agent to manage software companies' inbound sales.
• Check out the 13-slide pitch deck the startup used to raise funding.

An AI startup that helps software companies manage inbound sales has come out of stealth and secured $3.25 million in seed funding.

San Francisco-based Breakout, which was launched by ex-Googlers Sachin Gupta and Hitesh Aggarwal, has created an AI sales representative that can show demos and handle early-stage customer interactions.

Gupta used the analogy of walking into an Apple store and waiting in line to speak to a human sales rep. Instead, an AI sales agent would "instantly" give you all the information you needed about an iPhone and "quickly get into your buying process."

One big problem with some AI sales agents is that they can often provide scripted interactions instead of personalized responses. Breakout says its tools can answer questions and offer tailored interactions depending on each customer's needs.

"Especially as companies go through the growth stage and have a high-velocity sales process, they start facing a lot of challenges around the buyer experience," Gupta told Business Insider. "That's the problem we're trying to solve."

The startup makes its money by offering a usage-based pricing package, typically starting with 300 conversations a month and scaling up from there. Breakout says it currently has five key clients.

While there is still a lot of investor appetite for AI startups, Gupta said there was equally as much noise. That makes it "an interesting position for investors," he said, "because they want to deploy capital, but every day they also get to know another company which is building in the AI space."

He added, "Investors tend to have certain business objections. They tend to ask, do you have access to data? And in our case, we don't have a dependency on proprietary data."

Breakout's $3.25 million seed round was led by led by Village Global, with participation from Recall Capital and Z21 Ventures.

With the fresh funding, the startup plans to build out its tech capabilities so it can start serving its waitlist of customers.

Check out the 13-slide pitch deck, shared exclusively with BI, that it used to secure the fresh funds.

• Artificial intelligence has led to a boom in new startup creation and dealmaking.
• Business Insider researched startups that have strong founding teams and investor dollars.
• These are our top picks, listed alphabetically, of startups you could bet your career on in 2025.

After years of contraction, cost-cutting, and layoffs, there's been a resurgence of tech dealmaking in Silicon Valley thanks to the AI boom. Business Insider rounded up a number of technology and AI startups that are growing. Here are our top picks.

Abridge

HQ: Pittsburgh

Total raised: $462.5 million

What it does: The medical scribe startup translates patient-doctor interactions into clinical notes in electronic medical records.

What makes it promising: Abridge's business exploded in 2024 as investors rushed to fund companies automating administrative tasks in healthcare. The startup, which is backed by top VC firms, including Lightspeed Venture Partners and Bessemer Venture Partners, just raised $250 million in new funding at a million valuation. Its deals with top health systems, such as Kaiser Permanente, and its partnership with medical records giant Epic have made Abridge the healthcare AI startup to beat.

Anysphere

HQ: San Francisco

Total raised: $176 million

What it does: Anysphere makes AI coding software

What makes it promising: You may not have heard of Anysphere but you are likely familiar with its popular AI coding assistant, Cursor, that can predict a user's next line of code. The company recently raised $105 million at a $2.5 billion valuation. Notable investors include Benchmark, Andreessen Horowitz, and OpenAI.

Attention

HQ: New York

Total raised: $17.1 million

What it does: Attention uses natural language processing to fill out CRM programs and generate action items from sales calls.

What makes it promising: Some companies spend millions of dollars on customer relationship management programs, which are essentially software for sales teams that house crucial information about current and potential customers. The problem? Many teams don't properly fill out their CRM, rendering the investment useless. That's where Attention comes in — the startup uses AI to listen in on sales calls, fill out company CRMs with crucial information, and generate action items so a sales team member has the info they need to go back and close a deal.

Attention raised $14 million from Alven, Eniac, Frst, Liquid 2 Ventures, 645 Ventures, and Aglaé in October 2024.

Clasp

HQ: Boston

Total raised: $30 million, according to the company

What it does: Clasp helps employers secure critical talent before graduation, tackling workforce shortages in hard-to-hire fields like healthcare. Think of it like ROTC for critical professions. If a hospital system is facing a shrinking pipeline of nurses, it can partner with Clasp to access a national network of universities and training programs, match with current nursing students, and commit to repaying their student loans over a multi-year word period.

What makes it promising: Founded in 2018, Clasp has over 10,000 individuals on its platform and plenty of room to grow. While it's currently focused on building critical talent pools for the healthcare industry, the company plans to expand into other hard-to-hire industries. In 2024, Clasp raised over $10 million in a funding round led by Crosslink Capital and is actively investing in its growth team to scale employer and school partnerships.

CodaMetrix

HQ: Boston

Total raised: $95 million

What it does: CodaMetrix uses AI to analyze clinical notes and derive medical codes for billing and claims.

What makes it promising: Coding is a critical step of the revenue cycle management process for hospitals, typically requiring providers to manually assign numerical codes to medical services and diagnoses to ensure they get paid for their care.

CodaMetrix spun out of Mass General Brigham in 2019 to automate those administrative tasks and reduce provider burden, and it's captured a wave of investor interest in the sector, last raising a $40 million Series B round in March. The startup has worked with top health systems like Mayo Clinic and Yale Medicine to develop new revenue cycle management solutions, and it added some key hires to its executive team last year, including a new chief technology officer and COO.

Cohere Health

HQ: Boston

Total raised: $106 million

What it does: Cohere Health automates the pre-authorization process for medical treatments.

What makes it promising: Cohere Health works with health plans like Humana and Geisinger to make the prior authorization process more efficient and accurate, using AI to save money for the plans and reduce the number of unnecessary denials for patients. The startup last raised a $50 million Series B extension in February 2024, led by Deerfield Management and including existing investors, including Define Ventures and Flare Capital Partners.

Cohere has announced a number of new products in the last year, including tools released in January to help health plans meet prior authorization compliance standards set by the Centers for Medicare and Medicaid Services.

Coram AI

HQ: Sunnyvale, California

Total raised: $30 million

What it does: Coram AI puts agentic AI software into existing security systems and cameras.

What makes it promising: The US is filled with businesses and buildings that have non-operational security systems, Coram says. The startup's solution is an AI software that ports onto existing security hardware systems to provide generative AI visual security via AI agents that can identify and track threats in real time.

Coram raised $13.8 million in January from Battery Ventures, 8VC, and Mosaic Ventures.

Cortica

HQ: San Diego

Total raised: Over $300 million

What it does: Cortica provides virtual and in-person pediatric care for autism, as well as commonly co-occurring conditions like behavioral issues and sleep disorders.

What makes it promising: Cortica has set itself apart by going after value-based care contracts with health plans and employers that pay the startup for better patient outcomes, a rarity in specialized mental healthcare. The startup employs more than 2,000 providers that help care for children with autism at its clinics, in the home, or virtually, aiming to deliver "whole-child care" through everything from physical therapy to speech-language pathology to neurology. Cortica most recently raised an $80 million round in November, co-led by JP Morgan's healthcare investment fund Morgan Health and Nexus NeuroTech Ventures.

Daedalus

HQ: Karlsruhe, Germany

Total raised: $32.6 million

What it does: Daedalus helps factories and their production robots operate more efficiently.

What makes it promising: Launched by ex-OpenAI engineer Jonas Schneider, who was a key part of the AI juggernaut's robotics team, Daedalus was part of Y Combinator's winter 2020 cohort. The startup, which also has an office in San Francisco, uses AI robotics technology to cull the need to reprogram production robots constantly. Instead, it automates a lot of the tasks associated with the manufacturing process; for example, if clients give Daedalus a CAD drawing, it will render a fully-completed version of the drawing.

In February 2024, the startup raised a fresh $21 million Series A. The funding will help Daedalus in its mission of automating the manufacturing process across various industries, from semiconductors to healthcare.

Decagon

HQ: San Francisco

Total raised: $100 million

What it does: Decagon develops AI support agents that autonomously resolve customer inquiries over chat, email, or voice calls.

What makes it promising: The company raised $100 million, including a $65 million Series B, late last year. Bain Capital Ventures led the Series B round, and Elad Gil, A*, Accel, Bond Capital, and Acme Capital participated. According to the company's blog, the fundraise quadrupled Decagon's valuation. Bilt, Duolingo, Eventbrite, Notion, and Rippling use Decagon to manage interactions with customers by gathering data and reviewing conversations.

Decart

HQ: San Francisco

Total raised: $53 million, according to the company

What it does: Decart is an AI research lab focused on efficiency. Its infrastructure platform aims to dramatically cut the costs of training and running foundation models.

Why it's promising: Last fall, Decart emerged from stealth with $21 million in seed funding from Sequoia and Oren Zeev and launched a demo, Oasis, that captivated the tech world. Oasis's video platform enables users to create interactive, open-world experiences from a single uploaded image and generates content in real time based on user input. In December, the Israeli-founded startup secured an additional $32 million from Benchmark and other investors. Since then, Decart has doubled its team size and continues developing new products.

Elise AI

HQ: New York

Total raised: $172 million

What it does: EliseAI sells AI assistants, primarily to housing operators, as well as healthcare providers. These speed up menial tasks such as maintenance requests and scheduling appointments.

What makes it promising: The startup hit a unicorn valuation in 2024 with a $75 million Series D. Its technology is revolutionary for the housing sector, which previously suffered from inefficient technology, resulting in consumers absorbing additional costs, said founder and CEO Minna Song.

When arranging house viewings and meetings, keeping up with messaging prospective tenants can take up a lot of time and energy. Elise AI's chatbot automates these interactions so they free up time for management teams to pursue more meaningful work. The tech has also been embraced by the healthcare industry, which experiences similar pain points for managing invoices and bills, as well as patient appointments.

Flo Health

HQ: London, United Kingdom

Total raised: $300 million

What it does: Flo is a digital women's health company, which provides a period tracking and wellness app.

What makes it promising: Flo became the first digital women's health company to hit a unicorn valuation in 2024, following a $200 million raise from General Atlantic.

The startup, which launched in 2015, ballooned in popularity as it offered a comprehensive suite of products, such as period tracking and personalized insights into reproductive health via its app. After Roe v. Wade was repealed, Flo developed an 'Anonymous Mode' setting that would allow users to access the app without associating any identifying information with their health data.

Following its fundraise in 2024, the startup is making a big hiring push in Lithuania — recruiting for over 100 roles in Vilnius. It will also expand its user base and double down on offerings for those with menopause.

Glean

HQ: San Francisco

Total raised: $560 million, according to the company

What it does: Glean makes search chatbots and agents for businesses, allowing workers to search for information across various systems and applications and create and summarize content.

What makes it promising: Glean's business is taking off as organizations seek quick productivity gains. Founded by a group of former Google Search engineers, Glean topped $100 million in annual recurring revenue last fiscal year, up from $50 million earlier in the year. The company plans to expand into new markets and verticals in 2025 to keep up its momentum.

Grow Therapy

HQ: New York City

Total raised: $178 million

What it does: Grow Therapy works with independent therapy practices to streamline their administrative tasks and connect patients with therapists covered by their insurance.

What makes it promising: Grow helps therapists start and run their own mental health practices. SignalFire founder and CEO Chris Farmer named Grow Therapy to Business Insider's list of the most promising healthcare AI startups of 2024, citing the startup's focus on handling administrative tasks "so therapists can focus on their patients and control their own schedule instead of being underworked and underpaid at someone else's practice."

Grow Therapy most recently raised $88 million in Series C funding in April, led by Sequoia Capital.

Harvey

HQ: San Francisco

Total raised: $500 million

What it does: Harvey is a developer of a generative AI legal tech platform for lawyers and paralegals to help with contract analysis, due diligence, litigation, and regulatory compliance.

What makes it promising: Many startups are attempting the herculean task of disrupting the legal industry, but Harvey is in the pole position. Backed by big-name investors like Sequoia and Kleiner Perkins, Harvey has ridden the AI wave to recently double its valuation to $3 billion in a fresh $300 million round of funding. In 2024, Harvey saw 4x annual recurring revenue growth and now has 235 customers in 42 countries.

Hue

HQ: Remote

Total raised: $4.5 million

What it does: Hue helps brands and retailers sell online by collecting user-generated video reviews and embedding that content into product pages.

What makes it promising: Hue is bringing the power of TikTok-style video reviews to brands and retailers, significantly increasing conversion rates and time spent on-site. Founded by a trio of women who come from the consumer industry they now serve, Hue closed on $4.5 million in seed funding last year from Fika Ventures, Underscore VC, and others.

Knime

HQ: Zurich, Switzerland

Total raised: $53.8 million

What it does: Knime has built a low-code, open-source data analytics platform for businesses.

What makes it promising: The startup is headquartered in Switzerland but has a global presence, with offices in Texas and Berlin. Its mission is to democratize data analytics and utilize generative AI to make that mission more accessible, cofounder and CEO Michael Berthold previously told Business Insider.

The startup raised $30 million in equity funding from Invus in August 2024 and serves over 400 enterprise customers — including the likes of Audi, Novartis, and P&G.

Landbase

HQ: San Francisco

Total raised: $12.5 million

What it does: Landbase uses AI agents to automate businesses' go-to-market procedures.

What makes it promising: Launched in 2023, Landbase has quickly applied the use of agentic AI to automating GTM strategies, training its GTM Omni model on billions of data points.

In July 2024, it raised a fresh $12.5 million from First Minute Capital and 8VC. It also recently acquired LavaReach, an AI-powered prospect research tool.

Legora, formerly Leya

HQ: Stockholm with offices in London

Total raised: $37 million

What it does: Lawyers use Legora to streamline legal work across reviewing, drafting, and research.

What makes it promising: Just months after graduating from the storied startup accelerator Y Combinator, Legora raised back-to-back rounds of funding from investors like Benchmark, Redpoint, and Jack Altman's fund Alt Capital. The company has so far grown its business in Europe and the US and is now quickly expanding to new markets. The website's careers page shows the company is hiring go-to-market managers in New York, Madrid, and London.

Midi Health

HQ: San Francisco

Total raised: $100 million

What it does: Midi partners with employers and health systems to provide virtual care for menopause.

What makes it promising: Midi is leading a growing market for menopause support as women's health investors expand their reach beyond fertility and maternal care. 18% of employers surveyed by Mercer said they plan to provide menopause benefits to employees in 2025, up from a measly 4% in 2023. Midi provides virtual services, including hormonal-replacement therapy and lifestyle support for those struggling with hormonal changes as they age, navigating symptoms like hot flashes and weight gain through perimenopause and menopause.

Midi also works with health systems to offer specialized telehealth services and coordinate care alongside a patient's in-person doctors. The startup raised a $63 million Series B round last year from dozens of angel investors, including actress Amy Schumer and former Meta COO Sheryl Sandberg, as well as VC firms like GV (Google Ventures) and Emerson Collective.

Neubird

HQ: Redwood City

Total raised: $44.5 million

What it does: Uses artificial intelligence to monitor, analyze, and resolve IT issues for companies.

What makes it promising: Hawkeye, the startup's AI-powered ITOps engineer, automates the detection and resolution of IT issues, freeing software engineers from routine troubleshooting. The startup's growing customer base includes both startups and large financial institutions, according to TechCrunch. The startup raised a $22.5 million seed extension round led by Microsoft's M12 in December, just eight months after raising a $22 million seed round from Mayfield, TechCrunch reported.

Nimble

HQ: San Francisco

Total raised: $221 million

What it does: Nimble develops fully autonomous e-commerce fulfillment centers powered by its warehouse robots that can retrieve inventory, pick items, pack orders, and sort packages.

What makes it promising: Backed by FedEd and Accel, Nimble is building a national network of next-generation robotic warehouses to provide faster, lower-cost logistics. It aims to solve a critical pain point for customers like Puma and AdoreMe, who are attempting to scale operations while facing a national shortage of warehouse workers. The company most recently raised $106 million in a round co-led by FedEx and Cedar Pine that propelled its valuation to $1 billion.

Norm Ai

HQ: New York

Total raised: $38 million

What it does: Builds AI agents to automate compliance tasks and regulatory assessments.

What makes it promising: Norm's AI platform takes complex regulations and converts them into code that can be parsed by computers, allowing companies to clearly explain compliance findings, for example. The startup raised three rounds of funding — a Series A and two follow-on investments — in just 11 months. Coatue Management led its $27 million Series A, and Bain Capital Ventures, Blackstone Innovations Investments, and others participated.

Please, formerly MultiOn

HQ: Palo Alto, California

Total raised: Undisclosed

What it does: Please develops an AI assistant that helps consumers with their plans, using agents to complete actions like booking trips and managing reservations.

What makes it promising: The startup, which rebranded from MultiOn to Please in January, develops web-based AI agents that are powered by LLMs. Major players like Amazon and General Catalyst invested in the company in a round that valued it at $100 million, The Information reported.

Reality Defender

HQ: New York

Total raised: $40 million

What it does: Reality Defender has developed a deepfake detection platform that spots AI-generated content.

What makes it promising: As the use of AI-generated content burgeons, the technology has also been increasingly used to create fraudulent content and misinformation. Reality Defender's platform can detect if something is AI-generated in images, text, video, and audio.

In particular, the startup has found a niche in providing its services to enterprise clients to help identify deepfakes. It has developed an API and web app that allows users to analyze content and gauge if it's been modified by AI; however, it doesn't directly discern if something is a deepfake. Rather, users are given inference points so they can determine the extent to which AI has altered something.

In October 2024, the startup raised a fresh $33 million to grow its offerings in the financial sector.

Remark

HQ: Boston

Total raised: $10 million

What it does: Remark develops a shopping guidance platform that connects shoppers with online product experts.

What makes it promising: Remark helps shoppers make purchase decisions by allowing them to asynchronously chat with product experts, both human and AI, simulating the experience of chatting with a sales associate at a brick-and-mortar store. The two-year-old company helps consumers looking to purchase items in the fashion, home goods, outdoor, baby products, beauty, and skincare industries, Remark told Business Insider. And it's already seeing promising results: Brands using Remark have seen a 10-12% revenue lift and a 30+% conversion rate, according to the company.

Robin AI

HQ: London and New York

Total raised: $71 million

What it does: Robin AI develops an AI legal assistant that drafts and analyzes contracts for companies and their legal teams.

What makes it promising: Robin AI announced not one but two rounds of funding in 2024: a $26 million Series B, led by Temasek, and a $25 million follow-on investment, with participation from Paypal Ventures and Cambridge University. The company's AI-powered platform helps in-house counsel teams and enterprises streamline their contract review processes. Richard Robinson, who worked as a lawyer at Clifford Chance, and James Clough, previously a machine learning researcher, founded the company in 2019.

Rogo

HQ: New York

Total raised: $27 million

What it does: Rogo develops an AI agent that helps Wall Street professionals with tasks such as company research and memo drafting.

What makes it promising: Investment banking may look high-octane on HBO's "Industry," but working on Wall Street is a grind. Enter Rogo. The AI-powered platform helps analysts quickly analyze earnings, construct market maps, and other tasks. Two of Rogo's cofounders, Gabe Stengel and John Willett, previously worked in investment banking. Investors include Khosla Ventures, Jack Altman's AltCapital, AlleyCorp, and BoxGroup.

Rox

HQ: San Francisco

Total raised: $50 million

What it does: Rox's team of AI sales assistants automates tasks and provides data-driven insights for sales teams.

What makes it promising: AI-powered tools like Rox are gaining traction with sales teams by reducing administrative work and improving deal execution. The company streamlines CRM updates, summarizes relevant news events, and drafts outreach in its platform, helping sales reps focus on closing deals rather than on tedious tasks. As of November 2024, over 35 enterprise sales teams from companies such as MongoDB and Ramp have used Rox. The startup raised both its seed round, led by Sequoia with participation from Google Ventures, and its Series A, led by General Catalyst, in stealth. It's currently in public beta.

Sierra

HQ: San Francisco

Total raised: $285 million

What it does: Seirra's AI-powered conversational agents interact with customers.

What makes it promising: Founded by OpenAI chairman and ex-Salesforce co-CEO Bret Taylor and former Google executive Clay Bavor, Sierra's valuation soared to $4.5 billion at the end of 2024. Just don't call it a chatbot, as Taylor prefers to be thought of as "conversational AI." Whatever you call it, companies like ADT, Casper, and Sonos have used Sierra to handle customer service inquiries.

Skyfire

HQ: San Francisco

Total raised: $8.5 million

What it does: Skyfire is a payment network that lets AI agents autonomously spend money on behalf of their human counterparts.

What makes it promising: With AI agents expected to be a big theme in 2025, investors are excited about the types of tasks they can take over from humans. While other AI agents are handling customer service and sales calls, Skyfire is an early agentic player in the fintech space and is tackling the regulatory and societal considerations that come with giving a robot license to swipe a credit card.

Skyfire launched from stealth last summer and raised $8.5 million in seed funding from financial firms Neuberger Berman, DRW, and Brevan Howard Digital, plus Intersection Growth Partners, Arrington Capital, RedBeard Ventures, and others.

Smartcat

HQ: Amsterdam

Total raised: $75 million

What it does: Smartcat provides AI-generated translation services for businesses.

What makes it promising: For companies that want to scale globally, Smartcat offers a more cost-effective solution than hiring a gaggle of human translators. Smartcat's AI can translate both written and spoken words into more than 280 languages, making it easier to deploy corporate content, such as marketing materials and internal training videos, to office locations around the world.

Smartcat raised a $43 million Series C last year from Left Lane Capital, Koro Capital, Marbruck Investments, and Chrome Capital.

StackGen

HQ: San Francisco

Total raised: $12.3 million

What it does: StackGen uses AI to auto-generate infrastructure such as servers, databases, and networking from code.

What makes it promising: The AI revolution is coming for developers, with plenty of startups cropping up to help — and in some cases, replace — software engineers designing apps and building websites. StackGen is unique because it operates at the infrastructure layer of software development: Its AI reads code created by human developers and uses the information to generate technical infrastructure like servers and databases. StackGen raised $12.3 million last fall from a group of investors, including Thomvest Ventures, FireBolt Ventures, WestWave Capital, and Secure Octane.

Sublime Security

HQ: Washington, DC

Total raised: $94 million

What it does: Sublime's email security platform detects and prevents malicious behaviors in the inbox, enabling organizations to defend against phishing, email fraud, and other cyberattacks.

What makes it promising: Sublime's business has exploded as generative AI gives attackers a way to rapidly produce mass spear-phishing campaigns. The company has quadrupled its customer base over the past year and added enterprise customers like Elastic, Benteler, and SentinelOne to a roster of existing customers like Spotify, Reddit, and Brex. The company has won backing from top investors, including IVP, Index Ventures, and Slow Ventures.

SuperAGI

HQ: Newark, Delaware

Total raised: $15 million

What it does: SupaerAGI develops AI Agents for fully automated sales, marketing, support, and app development.

What makes it promising: SuperAGI got a big boost last year, picking up funding from Newlands VC, the secretive firm started by WhatsApp cofounder Jan Koum. Aiming to supercharge business teams, SuperAGI is used by developers at Google, Tesla, OpenAI, and Microsoft.

Synthesia

HQ: London

Total raised: Over $350 million

What it does: Synthesia is an AI video creator that helps companies with tasks such as employee training, customer support, and sales.

What makes it promising: Founded in 2017, Synthesia was early to the generative AI boom. It reportedly doubled its valuation in 2024 and moved beyond video creation to help businesses solve a wider range of needs. More than 5,000 companies use Snythsia, from Heineken to Dupont to Zoom.

Together AI

HQ: San Francisco

Total raised: $232 million

What it does: Together AI has created an open-source generative AI and infrastructure platform for developing AI models. The company runs data centers suited specifically for AI workloads.

What makes it promising: The company most recently raised $106 million in a round led by Salesforce Ventures that saw its valuation cross the $1 billion mark. Other big-name investors include Coatue, Kleiner Perkins, NEA, Greycroft, and Nvidia. The startup is reportedly raising another round of funding that would value it at $3 billion.

Torq

HQ: New York

Total raised: $192 million

What it does: Torq has created autonomous security operations to help companies guard against cyber attacks.

What makes it promising: Torq achieved 300% revenue growth and increased its headcount by 200% in 2024, according to the company. It recently hired a new head of sales Usman Gulfaraz, to help the company get to $100 million in annual recurring revenue for 2026. Customers include Chipotle Mexican Grill, Inditex, PepsiCo, Procter & Gamble, and Siemens.

Unify

HQ: San Francisco

Total raised: $18.2 million

What it does: Unify is a developer of a performance management system for sales teams.

What makes it promising: Backed by OpenAI and Thrive Capital, Unify helps salespeople tailor "warm outbound" emails that are less likely to get lost in crowded email boxes. Unify's growing team includes ex-staffers from Spotify, Airbnb, and Ramp.

Vapi

HQ: San Francisco

Total raised: $20 million

What it does: Vapi is building an infrastructure tool for developers to build AI voice agents.

What makes it promising: Investors are foaming at the mouth to back promising AI agents, and one group of startups is specifically using the tech to understand spoken commands. One of these so-called AI voice agent startups is Vapi, which is creating a tool for developers to create, test, and deploy AI voice agents of their own that can be applied in a number of business settings, from reception desk to employee training to sales call.

Vapi raised $20 million at the end of 2024 from Bessemer Venture Partners. Abstract Ventures, AI Grant, Y Combinator, Saga Ventures, and Michael Ovitz.

Writer

HQ: San Francisco

Total raised: $326 million

What it does: Writer is a full-stack generative AI platform that gives businesses tools to create their own AI applications and automate other workflows.

What makes it promising: Writer has carved out a niche in enterprise AI and offers a secure, customizable generative AI platform tailored for businesses — which sets it apart from more generalist models like OpenAI. The company has attracted major clients, including Fortune 500 firms, by focusing on data privacy, compliance, and domain-specific AI solutions, and its recently released AI model emphasizes control, security, and enterprise-grade performance.

Writer raised a $200 million Series C in November 2024 from Premji Invest, Radical Ventures, ICONIQ Growth, Adobe Ventures, B Capital, Citi Ventures, IBM Ventures, Salesforce Ventures, Workday Ventures, Accenture, Balderton, Insight Partners, and Vanguard. The round valued the startup at $1.9 billion.

7AI

HQ: Boston

Total raised: $36 million

What it does: 7AI uses AI agents to autonomously respond to alerts and investigate cyber threats on behalf of security operations teams.

What makes it promising: 7AI cofounders Lior Div and Yonatan Striem Amit previously cofounded Cyberreason, another cybersecurity company. 7AI raised a $36 million seed round in June 2024 that valued the company at over $100 million. The financing was led by Greylock Partners, with participation from Spark Capital and CRV.

• Mistral's CEO, Arthur Mensch, told BI that DeepSeek is a "great moment for open-source models."
• Mistral, a Paris-based startup that uses open-source, is seen as Europe's answer to OpenAI.
• Mensch said Mistral is focusing on agentic AI with a revamped version of its Le Chat app.

While markets panicked about DeepSeek, the CEO of Mistral welcomed the Chinese startup's new models as a boost for the open-source world.

Since its launch in 2023, Paris-based AI startup Mistral has advocated for open-source, which, broadly, is software that anyone can use or modify.

"We like to think of DeepSeek as the Mistral of China," Arthur Mensch told Business Insider in an interview. "We think that this is a great moment for open-source models, and companies like this — Mistral and DeepSeek — have participated in that and built on top of each other."

Like DeepSeek, many of Mistral's large language models are open-weight, which means they primarily share the model parameters rather than the full code base.

Mensch is one of many in the tech industry who point to the benefits of open source, such as faster development cycles and no extensive licensing fees.

"We've always been believers in open-source prevailing because of the flywheel, because of everybody building on top of each other, because of the fact that it just makes things more efficient," Mensch told BI.

Last month, investors questioned the scale of large AI infrastructure investments after DeepSeek released R1 models that claimed to match the performance of rival models like ChatGPT — but at a fraction of the cost.

In addition to raising questions about efficiency, it also prompted a rethink in the AI world about open versus closed source.

In a Reddit AMA session earlier this month, OpenAI boss Sam Altman admitted to being "on the wrong side of history" regarding his stance on open-source AI. He added that the company may need to figure out a new strategy instead of its closed-source approach.

Mensch said Mistral, which is seen as Europe's answer to OpenAI, has always prioritized efficiency — including when it comes to capital.

With $1 billion in funding from the likes of Andreessen Horowitz, Lightspeed, and General Catalyst, Mistral takes the mantle of Europe's most-capitalized AI startup.

"It doesn't take $100 billion to make enterprises adopt a technology and adapt it to their use cases," he said. "We're a very well-capitalized company, but we're not spending hundreds of billions."

Mistral's agentic AI and IPO plans

In addition to touting its open-source ambitions, Mensch told BI that Mistral would expand its capabilities in agentic AI, a type of software that can perform tasks autonomously.

The startup is doing so with a revamped version of Le Chat, which launched as an app for the first time last week.

"Le Chat is a place where you can build your agents, and you should see this as a place where every employee and consumer can create automation and spawn agents," he added. The best aspect of AI agents is that non-developers can also use them," Mensch said.

The startup also has goals to reduce its dependence on Big Tech; before the Paris Action Summit this week, it announced that it would invest billions of euros in building its own data center in France.

"It's a choice we are making to have control over the whole value chain, from the machine to the software," Mensch told French TV channel TFI.

Mistral is also doubling down on its presence outside its home country. It's expanding in the US, a market where it's seeing a lot more commercial traction, and it has opened a Singapore office in light of a burgeoning user base in southeast Asia.

While Mensch previously indicated that Mistral would eventually go public, he clarified to BI that the company has no immediate plans to pursue that route.

"I said that we were an independent company, and as a successful independent company, the natural path would obviously be IPO long term — but this obviously doesn't mean that we're preparing anything for an IPO for now," he said.

• EnCharge AI has secured $100 million to advance its energy-efficient AI chip technology.
• The startup says its in-memory computing chip uses less energy to run AI calculations.
• Check out the pitch deck it used to secure the fresh funding.

A startup developing efficient AI chips has secured $100 million in a funding round led by investment firm Tiger Global.

California-based EnCharge AI, founded in 2022, has created an in-memory computing AI chip that it claims provides a twenty-fold improvement in energy efficiency.

This allows inference computation — when AI makes predictions based on new data — to take place outside the cloud and on local devices. As a result, advanced AI applications wouldn't always have to be powered by data centers and could instead be performed on smaller computers and edge devices.

The startup's product is a commercial expansion of research from its CEO, Naveen Verma, which explored non-volatile memory devices — a type of memory that holds stored information even when its source of power is removed. This, in theory, makes devices more energy-efficient as they don't need to consume power in order to retain information.

Other computing architectures don't use in-memory techniques, which can make devices unsuitable for running AI calculations because it takes too much time and energy to move data back and forth between memory and processing units. EnCharge's in-memory chips use static random-access memory, or SRAM, to improve compute efficiency.

The hardware can be deployed for everything from automotive sensing to smart retail and industrial robotics.

The Series B was led by Tiger Global, with participation from Maverick Silicon, Capital TEN, SIP Global Partners, Zero Infinity Partners, CTBC VC, Vanderbilt University, Morgan Creek Digital, and others.

Previous investors backing the round include RTX Ventures and Anzu Partners, among others. It brings the startup's total funding to around $144 million.

EnCharge is also working with semiconductor giant TSMC to refine its first-generation commercial chips.

With the fresh funding, the startup said it would bring its AI accelerator solutions to market.

Check out the pitch deck it used to secure the fresh funding.

• AI companies face high compute costs and energy demands while developing and deploying models.
• A crop of startups are building solutions to make AI more energy and cost-efficient.
• From data center cooling to chip efficiency, VCs say these startups are making AI cheaper and greener.

DeepSeek sent a clear message to Silicon Valley startups: It's possible to do more with less.

Now, competitors in the US are scrambling to replicate the Chinese startup's approach, which appears to rival the performance of top AI models in the US but seemingly at a fraction of the cost.

That's prompted industry insiders to question the billions of dollars being spent on AI infrastructure. It has also spotlighted startups that are developing solutions to lower the high costs of developing, deploying, and running AI models.

Training AI requires huge processing power, which is fueled by clusters of graphics processing units, or GPUS. They consume a lot of energy, and these power circuits are largely provided by data centers.

"The energy consumption of training a large AI model can produce emissions equivalent to the lifetime emissions of multiple cars," Andreas Riegler, general partner at APEX Ventures, told Business Insider. "As models grow in size, the demand for energy scales exponentially, raising sustainability concerns for future applications," he added.

There are many approaches toward making AI greener and cheaper to use, Riegler told BI. Startups can improve software efficiency, develop more energy-efficient chips, and tap into renewable energy sources, he said.

BI spoke to seven investors based in Europe and the US, asking them to put forward startups that are helping make AI cheaper and greener to use.

Mobius Labs

Total raised: $7 million

Backers: Atlantic Labs, APEX Ventures, Lunar, and angels

Suggested by: Andreas Riegler, founding partner at APEX Ventures

What it does and why it's one to watch: Berlin-based Mobius Labs has developed an AI platform that claims to reduce AI costs by using 10 times less computing power.

While it launched in 2020, the startup pioneered a new product line in 2024 — which allows it to "transition from a computer vision recognition company to a full multimodal generative AI provider," said the startup's cofounder and CEO, Appu Shaji.

Mobius Labs' platform aims to reduce the energy required to run AI models, which is otherwise costly and energy-intensive.

The startup says it has "quantized" Meta's Llama70B model — essentially, divided it into smaller, more measurable increments — "without losing accuracy, and enabling it to run on a single GPU instead of four," said Riegler, whose VC firm has backed the company. "This method skips unnecessary calculations without sacrificing the accuracy" of the output, he added.

Gemesys

Total raised: $12.8 million

Backers: Atlantic Labs, NRW.Bank, and Plug and Play Tech Center

Suggested by: Andreas Riegler, founding partner at APEX Ventures

What it does and why it's one to watch: AI-hardware startup Gemesys has developed a chip that works similarly to the human brain and positions itself as a more energy-efficient alternative for AI applications.

Launched in 2023, the German startup has designed a piece of hardware called a neuromorphic chip. These differ from traditional chips, which process information in a "step-by-step manner," said Riegler.

"Neuromorphic chips are built to work like a network of neurons (the brain's nerve cells), which can handle multiple things at once and adapt over time," he told BI. "This makes them very good for tasks like pattern recognition, image processing, and sensory data analysis."

According to Riegler, this makes it a faster alternative to traditional chips because it powers AI applications that need to run continuously — such as smart sensors, robots, or small devices — more efficiently.

Cohere

Total raised: $975 million

Backers: Thomvest Ventures, Index Ventures, Tiger Global, Radical Ventures

Suggested by: Umesh Padval, managing director at Thomvest Ventures

What it does and why it's one to watch: Cohere is a foundational model layer company that provides large language models to enterprises so they can develop and rapidly deploy AI applications.

Based in Toronto, Canada, the company allows enterprises to host Cohere's proprietary language model on its own servers. The startup has secured partnerships with the likes of McKinsey and software giant Oracle. In September 2023, it launched a chatbot called Coral.

"The company is providing high-performance LLM models, with the highest accuracy at the lowest cost — thus providing the lowest cost of ownership for its customers," said Padval, an investor in the company.

Cohere raised $500 million in July 2024 to grow its technical team and double down on its enterprise offerings.

Corintis

Total raised: $9 million

Backers: Blueyard Capital, Founderful, and Acequia Capital

Suggested by: Antonia Albert, principal at Founderful, and Rick Hao, partner at Speedinvest

What it does and why it's one to watch: Lausanne, Switzerland-based Corintis has developed a cooling solution for AI applications and data centers.

The startup has created a technology called a "precision microfluidic cooling solution," which aims to improve semiconductor chips' performance by preventing them from overheating.

Corintis embeds a network of microscopic channels directly into the chips. The cooling liquid extracts heat, achieving 10 times more efficiency than current solutions, said Albert.

Apheros

Total raised: $1.85 million

Backers: Founderful

Suggested by: Antonia Albert, principal at Founderful

What it does and why it's one to watch: ETH Zurich spinout Apheros launched in August 2023 with a patented technology that helps data centers cool down.

"We take a device called a heat sink, whose purpose is to bring heat away from your chip, your GPU, and into the surroundings and the coolant — and we've optimized it through the usage of a metal foam," cofounder and CEO Julia Carpenter previously told BI. This process allows heat to dissipate more easily in data centers, ultimately using up less energy in the process.

"By 2030, an estimated 6% of global energy consumption will be used for cooling data centers — and a need for cost-and energy efficient liquid-based solutions is inevitable," said Albert.

Syntiant

Total raised: $121 million

Backers: M12, Microsoft, Intel Capital, Alumni Ventures

Suggested by: Samir Kumar, general partner at Touring Capital

What it does and why it's one to watch: California-based Syntiant has developed a method to make semiconductor chips more efficient by "combining purpose-built silicon with an edge-optimized data platform," said Kumar.

The company says it has streamlined the conversion of raw and synthetic data into "quality machine learning models" — making data processing for AI more efficient.

Kumar told BI that Syntiant is focused on improving power and performance efficiency for inference — a technique in which AI makes predictions about new data — that takes place on battery-powered internet-connected devices. Improved efficiency on these devices could lead to more AI-enabled sensors for speech, sound, vision, and other modalities, he added.

DBtune

Total raised: $3 million

Backers: 42 Cap, Axeleo Capital, JVH Ventures, Tiny Supercomputer, Exxeta Ventures, Tola Capital

Suggested by: Sheila Gulati, managing director at Tola Capital

What it does and why it's one to watch: Sweden-based DBtune is a spinout from Stanford University. Launched in 2022, the startup aims to improve the way databases perform to reduce cloud costs for clients while lessening their environmental impact.

The startup uses AI and machine learning to optimize each database to the specific customer's workload and needs, which, it claims, makes database tuning more efficient.

"The company's core technology is a fully automated service designed to optimize database management, enabling businesses to achieve high database performance while minimizing their environmental impact," said Gulati. "Database efficiency is a game changer for a company's carbon footprints."

Cartesia

Total raised: $26.7 million

Backers: AWS Startups, 515 Ventures, Conviction Partners, Index Ventures, Lightspeed

Suggested by: Sheila Gulati, managing director at Tola Capital

What it does and why it's one to watch: San Francisco-based Cartesia, which was founded in 2023, is developing a new architecture for AI systems.

The startup has released a slate of new AI models that it says can run large models on smaller devices using less energy.

Cartesia "builds real-time, multimodal intelligence for every device using their pioneered State Space Models, enabling more efficient AI," Gulati told BI.

Mako

Total raised: $1.4 million

Backers: Flybridge, Jeff Dean, the Chief Scientist of Google Deepmind and head of Google Research

Suggested by: Chip Hazard, general partner at Flybridge Capital

What it does and why it's one to watch: Previously known as A2 Labs, Mako emerged from stealth in October 2024 with a $1.4 million pre-seed round.

The Boston-based startup aims to reduce compute costs by up to 70%. It has developed a technology that automates the process of GPU tuning — something that increases the processing power of a GPU — which is an otherwise manual and costly task.

Mako's platform generates a code that runs on any kind of hardware and optimizes GPU performance. This, in turn, ensures that AI models can run more efficiently at peak performance, the company website claims.

"Mako is a company to watch as they allow any AI model to be run on any hardware or cloud platform, and brings hardware independence to AI application developer," said Hazard.

PoroTech

Total raised: $26.1 million

Backers: Speedinvest, Cambridge Enterprise, IQ Capital Partners

Suggested by: Rick Hao, partner at Speedinvest

What it does and why it's one to watch: University of Cambridge spinout PoroTech, which launched in 2018, is developing a technology that makes semiconductor chips more energy-efficient.

It works with gallium nitride (GaN), a semiconductor material that is a better conductor than silicon. The team researched how to engineer GaN so that it can be applied efficiently to semiconductor-based applications.

"The company's groundbreaking work in GaN technology has positioned it at the forefront of innovation, particularly in the development of optical interconnects and optical processors that enhance energy efficiency, communication, and computation at the speed of light," Hao told BI. "This saves a ton of energy for data centers, etc."

Liquid AI

Total raised: $296 million

Backers: AMD Ventures, Bold Capital Partners, Capgemini

Suggested by: Chip Hazard, Flybridge Capital

What it does and why it's one to watch: MIT spinout Liquid AI emerged from stealth in October 2023.

The startup is working to build a general-purpose AI system called a liquid neural network. The aim is for this to use less compute power and energy while operating more transparently than the technology that powers mainstream chatbots and image generators.

Liquid neural networks are a type of architecture that processes data more efficiently than traditional neural networks, as they consistently adapt to new inputs.

The startup's "Liquid Foundation Models" can be applied to everything from detecting fraud to controlling self-driving cars.

"Unlike traditional transformer-based models, whose memory usage and inference time increase with longer input sequences, LFMs maintain near-constant inference time and memory complexity regardless of context length," said Hazard. "So they can process longer sequences without a significant increase in computational resources or energy consumption."

• Human.org has raised $7.3 million for safety tools to align AI with human values.
• The startup uses blockchain to verify identities and distinguish humans from AI agents.
• Backers in the funding round include HF0, Soma Capital, Spearhead, Pioneer Fund, and others.

A startup building a platform to ensure AI agents are aligned with human values and an app for people to verify they are human has raised $7.3 million.

San Francisco-based Human.org bills itself as an AI safety lab building "trust infrastructure" to keep AI accountable, said the startup's 23-year-old founder and CEO, Kirill Avery.

"We're two and a half years into AI-generated content, and it's becoming impossible to differentiate which content has been created by humans and which is not — but this is just the beginning," he told Business Insider.

With AI agents — software that can perform tasks autonomously —becoming more mainstream, Avery added that it's vital that human expression is preserved and that AI remains aligned with human intentions.

AI alignment is the process of ensuring an AI's actions are in step with the values of its human creators. The field gained mainstream traction as AI juggernauts raced to build high-performing models at breakneck speed, fuelling fears that the rise of a superhuman AI could disregard humanity's key values.

Human is developing two core products. Its "Human Network" platform, powered by a blockchain protocol, is an infrastructure layer for interactions between humans and verified AI agents. The startup's "Agent ID" protocol allows users to assign accountability to AI agents that can be traced back to their human creators. Because it's on the blockchain, it isn't controlled or owned by anyone and is accessible to the wider public.

Human's other key product is the Human App, which Avery says is "basically the solution to proving that you're a real human" and "helping people to show their humanity on different services."

The startup has created two forms of online ID — one for humans and one for AI agents — that act as a "digital passport," Avery told BI. "Signing in with Human ID is the essential part of knowing if this is a human or an agent you're talking to, as a new form of 'captcha'. The Human App is a place to manage your identity and all of your agent logs," he added.

The $7.3 million pre-seed was backed by HF0, Soma Capital, Spearhead, Pioneer Fund, Hummingbird VC, Brett Gibson, general partner at Initialized Capital, and others.

With the fresh funding, Human will continue to focus on its deep tech research.

• Mistral has launched a revamped version of its "le Chat" AI assistant.
• The tool can be used for consumer and enterprise use cases, CEO Arthur Mensch told BI.
• The French startup has been billed Europe's response to OpenAI.

Paris-based AI startup Mistral has launched a new version of its AI work and life assistant, "le Chat."

Mistral, one of Europe's most-funded AI startups, said le Chat is available on app stores for the first time from Thursday.

The tool lets users access news, upload documents, and track their projects. Most features are free, while the Pro tier costs $14.99 a month. Mistral said it plans to make le Chat available for enterprises with custom models and available through private business channels like Slack.

"There's also some elevated privacy as well where you can fully opt out of any data recording, which is very useful for professionals," cofounder and CEO Arthur Mensch told Business Insider in an interview.

Mensch uses le Chat to generate documents — and said it's a "useful sparring partner when it comes to thinking about strategy."

"Since Mistral is connected to all of our tools, it's great for me to understand what's going on because I can ask it to give me a summary of what's going on in the company every morning," he told BI.

Mistral said this iteration of le Chat, powered by its low-latency models, is its fastest version yet — providing an output of up to 1,000 words a second.

Users can generate photos, get personalized recommendations, and even build their own micro-apps on mobile, Mistral said.

Mensch said the company has seen the most user traction from financial services companies and the administrative sector in France.

"More recently, we've started to move forward with manufacturing companies, and we've announced a partnership with universities as well, where our technology will be made available to students and teachers in a way that will make their teaching or learning much easier," he added.

Mistral claims that this version of le Chat is much faster than its competitors' platforms. Unlike some of its rivals, Mistral's AI is open source, which the startup has touted since its launch. Its tool can also operate in multiple languages.

Investor appetite for AI assistants has soared amid a sustained interest in agentic AI — systems that can autonomously solve problems for humans.

Since its 2023 launch, Mistral has raised over $1 billion from investment heavyweights such as Lightspeed, Andreessen Horowitz, and General Catalyst.

However, its total investment is a drop in the ocean compared to its competitors in the US, such as OpenAI and Anthropic, which have collectively raked in tens of billions.

• Big Tech companies like Meta and Amazon have been scaling back their DEI programs.
• European tech industry insiders say the continent may follow the US' lead.
• It could result in Europe's tech firms refining DEI policies, said Hannah Leach, partner at Antler.

Big Tech is rolling back DEI measures. Europe's tech sector is yet to feel the same pressure, but industry insiders expect some of the sentiment from the US to spill over across the Atlantic.

Meta and Amazon have recently scaled back their diversity initiatives amid a wider pushback against DEI in corporate America.

Founders and investors in Europe see it as an omen of what could transpire on the continent.

"The US can be a lead indicator of what could happen in Europe concerning diversity," said Rodney Appiah, managing partner at Cornerstone VC, a fund that backs underrepresented founders. "We recognize that we are probably four to five years away from that sort of translating into Europe and certainly into the UK."

Appiah pointed to the collaborative nature of the tech and VC industry and the UK's dependency on the US tech market. "The US narrative will eventually have a contributing factor to the way in which Europe moves," he told Business Insider.

Big Tech inspiration

Diversity, equity, and inclusion programs in Europe could come under pressure if big-name companies in the US continue to scale back initiatives, said Michael Smets, professor at Oxford's Said Business School.

That's because "global companies have global policies," Smets told BI. "The idea that Microsoft et al. would roll back DEI programs in the US, but not in Europe, seems naïve."

While Microsoft has not announced any recent rollbacks of DEI initiatives, the tech giant laid off a diversity, equity, and inclusion team in July.

European tech companies would also look to their larger peers in the US for "inspiration," Smets added.

"When leaders don't know how to navigate an uncertain, new situation, they are likely to copy leading organizations in their field," he explained. "Doing the same as Big Tech — whether right or wrong — is much more defensible than the bold choice of going against the grain," he added.

Some US corporate heavyweights, such as JPMorgan, have said they remain committed to DEI efforts.

DEI differences

Europe's tech companies may look to Silicon Valley for inspiration, but they have historically had different approaches to DEI.

European founders and investors often cite the benefits DEI efforts can bring.

"The return on investment on DEI has been forgotten about as everything has become politicized," Amardeep Parmar, cofounder of BAE HQ, a platform for British-Asian entrepreneurs, told BI.

He added that hiring from underrepresented groups gives companies access to a wider talent pool. That can lead to commercial benefits, Parmar said.

"People aren't so easy to fool, though, and you've got to match your words with your actions — we returned 3-8x for our sponsor organizations in 2024," he said.

President Donald Trump has issued executive orders to scrap DEI offices and positions throughout the federal government and revoked a civil rights-era Equal Employment Opportunity executive order. Trump, along with Elon Musk, has instead called for an approach that prioritizes meritocracy.

Radha Vyas, cofounder and CEO of travel startup Flash Pack, told BI that DEI initiatives could "level the playing field for everyone" to ensure people are promoted on merit.

She pointed to data-driven processes "such as blind CV screening and merit-based promotions — that focus on skills and performance rather than subjective factors or unconscious prejudice."

Political buffer

Europe's current political system could act as a buffer against the influx of anti-DEI sentiment in the US.

"The UK political environment is quite different to several other countries because we're one of the very few Western economies that has a left-leaning government in situ, and so that provides a bit of a defense to kind of our politics descending into sort of culture wars," Appiah told BI.

Still, the rise of populist parties in Europe could supercharge a wave of negative sentiment against home-grown diversity policies.

Ruth Foxe Blader, general partner at Foxe Capital, pointed to the "rising legitimacy of populist movements" in Germany, France, and the UK.

"It's a way of doing that negative virtue signaling that politicians perceive people want to hear right now, and it's going to take a lot of courage for leaders of large companies in Europe to contest that sentiment," Blader told BI.

Some quarters of Europe's tech scene believe that any shift from existing diversity programs in Europe could be channeled into meaningful change.

"We're likely to see a move away from performative, tick-the-box DEI activity and a move towards DEI actions that are material to, and specifically designed for, and properly integrated within, the businesses in question," said Hannah Leach, partner at VC firm Antler.

She added that the US could also prompt more urgent conversations and reframing of what DEI initiatives do and what they should look like in Europe.

Industry insiders are expecting a DEI reckoning for Europe's tech sector — it's just a matter of when.

"When America sneezes, the rest of the world catches a cold," Appiah said.

• SoftBank is in talks to invest up to $25 billion in OpenAI, which could value it around $300 billion.
• The deal could make OpenAI the joint-second most valuable tech company after SpaceX.
• The funding round could see OpenAI pump $15 billion into the Stargate AI infrastructure project.

SoftBank is preparing to lead fresh investment into OpenAI at a $300 billion valuation, Business Insider understands — a move that would value the ChatGPT maker the same as TikTok owner ByteDance.

The Japanese investment giant is in talks to invest up to $25 billion in OpenAI, the Financial Times first reported.

While SpaceX is the most valuable private company in the world, the funding round could push OpenAI up to second spot alongside Chinese tech giant ByteDance. OpenAI currently ranks as the third-most valuable private tech firm.

Talks between OpenAI and SoftBank are ongoing, which means details around the the round size and valuation are subject to change. If the AI juggernaut hits its target valuation of $300 billion, it would nearly double its current valuation of $157 billion.

SpaceX's valuation is around $350 billion after the Elon Musk-owned company agreed to buy back $1.25 billion worth of stock at $185 a share in December. ByteDance, the parent company of TikTok, similarly catapulted its valuation to $300 billion following a buyback offer, offering investors a price of about $180 per share, the Wall Street Journal reported in November.

OpenAI could become one of three companies globally with a centibillion valuation; Stripe and Shein follow suit with valuations of $70 billion and $66 billion, respectively.

Prior to SoftBank's latest investment talks, OpenAI raised nearly $20 billion from investment heavyweights, including Thrive Capital, Khosla Ventures, and Nvidia.

If SoftBank invests $15 to $20 billion into OpenAI, it would overtake Microsoft — which has poured around $13 billion into the company so far — as OpenAI's lead investor.

Such a deal would mark the most significant bet yet on the generative AI boom from SoftBank and form a key part of the wider ambitions of its billionaire founder, Masayoshi Son, to usher in an era of "artificial super intelligence."

Son, who made his original fortune from a timely bet on Alibaba in the dot-com era, has previously spoken about his nonstop use of OpenAI's ChatGPT, and predicted AI that is 10,000 times smarter than humans will arrive by the midpoint of the next decade.

A fresh investment from SoftBank would also further deepen its relationship with OpenAI after the companies announced plans last week to form Stargate, a joint venture that aims to spend up to $500 billion on AI infrastructure projects in the US over the next four years.

SoftBank's Son will serve as chairman of the project, which was unveiled at the White House by President Donald Trump. Initial equity funders include Oracle and the UAE's MGX alongside SoftBank and OpenAI. The project will begin deploying $100 billion immediately, according to OpenAI.

It is not yet clear where the companies will source the capital from, with OpenAI currently lossmaking. It has been suggested that SoftBank's proposed equity investment could allow OpenAI to invest around $15 billion in Stargate.

SoftBank declined to comment. OpenAI did not immediately respond to a Business Insider request for comment.

• DeepSeek may have used OpenAI's models to train its own AI model, says a White House advisor.
• The Chinese startup's R1 model sparked market panic due to its efficient performance.
• OpenAI has also faced scrutiny over training practices and copyright lawsuits from media firms.

DeepSeek may have used OpenAI to create its Silicon Valley-shaking model, a top White House advisor said on Tuesday.

David Sacks, the White House's artificial intelligence and crypto czar, told Fox News "it's possible" that the Chinese AI startup engaged in intellectual property theft from OpenAI.

Sacks highlighted an AI training technique called distillation, in which a company uses information from an existing AI model to create a new model. Here, the bigger, more complex model — which is considered to be the "teacher model" — transfers much of its learnings to the smaller "student model."

"They can essentially mimic the reasoning process that they learn from the parent model and they can suck the knowledge out of the parent model," Sacks said.

DeepSeek's R1 model appears to match many of the capabilities of rival models but with much less compute — the expensive component powering AI. That sparked a stock market panic earlier this week, wiping hundreds of billions of dollars in value from companies selling AI chips, such as Nvidia.

"There's substantial evidence that what DeepSeek did here is they distilled the knowledge out of OpenAI's models. I don't think OpenAI is very happy about this," Sacks said, without citing specific evidence.

He said companies would take steps to prevent distillation.

"That would definitely slow down some of these copycat models," said the former venture capital investor and tech executive.

OpenAI has also faced scrutiny for its training practices. The company is facing several lawsuits over copyright infringement, including from media organizations like The New York Times.

OpenAI has its own model distillation offering. This allows smaller models to create their own datasets for distillation using OpenAI's API.

However, the company's terms of service prohibit users from "copying" any of its services or using its models to build rival ones.

Microsoft and OpenAI did not respond to requests for comment. DeepSeek did not respond to a request for comment, made on the Lunar New Year.

• Voice AI startups raised over $398 million in VC funding in 2024, per PitchBook data.
• The technology is expanding into enterprise uses like customer service and assistants.
• BI spoke to investors about the untapped opportunities in the nascent voice AI space.

Voice is fast becoming the new AI battleground.

From buzzy virtual assistants to speech synthesis tools, the technology has taken off in the past year.

While AI voice tech isn't new, the tools have rapidly become more sophisticated, driving adoption from the call center to recruitment agencies.

Its use cases are vast, from real-time audio transcriptions to generating synthetic voices from text prompts.

Investors looking for the next opportunity in the highly competitive AI market have thrown their checkbooks behind startups. According to PitchBook data, startups developing voice AI technology raised over $398 million in VC funding in 2024.

London-based PolyAI, which has developed voice assistants for call centers, secured $50 million in a funding round from Hedosophia. London and New York-headquartered ElevenLabs, which has developed a voice cloning technology, raised $100 million in January 2024 — and is said to be raising a further $200 million, Business Insider first reported.

"Recent breakthroughs in real-time speech-to-speech processing have unlocked new use cases, including virtual assistants, customer support, and voice-based productivity," said Sivesh Sukumar, an investor at VC firm Balderton. "Companies like ElevenLabs and OpenAI are at the forefront of this space, with ElevenLabs releasing a real-time API expected to drive further adoption."

Voice AI is a comparatively nascent space, so there isn't an established incumbent yet — but it's triggering investor excitement for the untapped opportunities in the sector, Sukumar added.

An expanding ecosystem

Startups are quickly identifying how to tailor voice technology to a host of enterprise and consumer needs. And with agentic AI a hot topic for CEOs, its overlap with voice technology could hold high potential.

PlayAI, a startup that is developing an AI platform for text-to-speech models and AI voice agents, raised $21 million in seed funding in November.

"We've seen a massive increase in interest in building voice agents, which a human can speak to just like it's another human," said PlayAI cofounder Hammad Syed. "Voice AI is going mainstream and will be a key interface in how people interact with technology. Investors definitely realize this opportunity," he added.

VCs scouring the ecosystem to make their next big bet are now looking at voice AI as a technology stack, said Steve Jang, founder and managing partner at Kindred Ventures, which also backed PlayAI. The firm's investment thesis is to back startups "with multiple layers with many use cases in consumer, enterprise, and creativity."

"First, there are specialized and foundational models. Second, there are infrastructure services and tools, which offer access and integration with these models. And then, perhaps most importantly, there is the vast vertical application space," he told BI.

The sector is also attractive to investors because voice is an easy category to cash in on. "You can price it by the outcome, so it's pretty easy to monetize," said Jonathan Userovici, general partner at VC firm Headline. "That's why you have so much revenue traction — it's pretty easy to get a return on investment, especially if you're replacing a human doing that labor."

Consumer appetite for voice AI has also skyrocketed. With more users preferring to take in information through audio formats like podcasts, Sukumar highlighted the growing consumer demand for voice control and audio interfaces. He built PersuAIsion, a voice AI platform that allows users to practice real-world conversations — from job interviews to first dates — because he saw the scope for voice to cater to such consumer needs.

"If OpenAI can capture the consumer voice agent, they'll be what Siri was supposed to be," he said. "I think there's going to be a lot more interfacing with personal devices, and there's just going to be better e-commerce consumer experience on that front."

Frontier labs are catching up

Despite its growing popularity, voice AI doesn't seem to have one established juggernaut just yet. Part of the reason could be that frontier labs have largely stayed away from the space, possibly due to a concern that a misuse of voice generation capabilities could result in a potential backlash, according to Air Street Capital's 2024 State of AI report.

"Despite scraping huge quantities of audio and video data, frontier labs have been slow to release text-to-speech products," said Nathan Benaich, founder and general partner of Air Street Capital. He pointed to OpenAI's Advanced Voice Mode, which was repeatedly postponed, and Google's NotebookLM, which "is relatively locked down."

AI experts had sounded the alarm about the possible rise of deepfakes in a year marked by global elections — but that didn't end up being the case.

"In all likelihood, labs were keen to avoid being dragged into the panics about deepfakes that often accompany major elections. I think it's inevitable they will play more in this space, just because the potential commercial opportunity is so large," Benaich said.

Big Tech may be slowly pivoting toward the trend. Amazon's plans to ramp up its voice assistant offerings through Alexa were delayed until 2025, and Apple recently bulked up its Siri feature by adding ChatGPT capabilities.

Still, Benaich noted that it won't be an easy task for any one company to take the crown. "Displacing companies like ElevenLabs, who already enjoy widespread adoption and have been optimizing their tools for enterprise users for years now, could prove challenging," he said.

• Uber's CEO says society needs to agree on an acceptable error-level from AI in the physical world.
• Dara Khosrowshahi, speaking at Davos, asked how much better AI would need to be than humans.
• Google's investment chief, Ruth Porat, said that society is more forgiving toward human errors.

Society will have to weigh up the acceptable level of errors from AI in real-world systems like robotaxis versus its potential benefits, Uber's CEO Dara Khosrowshahi has said.

"Part of humanity is its flaws, and we accept humans are going to make mistakes," he said, speaking during a panel at the World Economic Forum on Thursday. "I think one of the questions in certain AI applications — defense may be one of them, certainly, AI in the physical world — is, how much better does that AI have to be than a human being?"

Khosrowshahi pointed to the safety record of autonomous vehicle giant Waymo, a subsidiary of Google owner Alphabet, as an example. Google's chief investment officer, Ruth Porat, responded that while Waymo's technology was "meaningfully safer" than that operated by human beings, "there is more forgiveness when it's a human" making a mistake.

A report produced by Waymo in partnership with the Swiss Reinsurance Company said that its fleet of autonomous cars was safer than human drivers. The report, published in December, said that Waymo's cars faced 90% fewer insurance claims related to bodily harm.

Still, self-driving cars have faced backlash amid some high-profile safety incidents. Last year, Waymo issued a recall after two of its cars crashed into a pickup truck, while Tesla's Full Self-Driving system has been involved in multiple incidents.

Khosrowshahi asked how society should weigh up the mistakes of machines and algorithms "versus the benefits of AI coming into the forefront."

The discussion comes as the AI industry has continued to gather pace while grappling with issues around accuracy.

Apple recently came under scrutiny for sending users misleading information and summaries of sensitive news stories. Similarly, Google was under fire when its search tool gave users misinformed advice, such as telling them to put glue on their pizza, and eat one rock daily.

• Regeneration.VC is taking over the management of Sky Ocean Ventures' climate portfolio.
• The US-headquartered fund now counts nearly 50 companies in its global portfolio.
• Regeneration.VC is also expanding in Europe amid US policy shifts, adding a new partner.

Climate tech fund Regeneration.VC, which is backed by Leonardo DiCaprio as a limited partner, has taken over the management of Sky Ocean Ventures' climate portfolio.

Launched in 2022, Regeneration.VC bills itself as a consumer climate fund backing companies that are making industries such as food, apparel, and electronics more sustainable, partner Jamie Rowles told Business Insider in an interview.

Its portfolio includes the likes of recycling startup Greyparrot and textile dyeing company Colorifix.

Sky Ocean Ventures, the impact investing arm of British telecoms company Sky, had backed early-stage climate and cleantech startups working on everything from recycling technology to combatting the plastic pollution crisis. Regeneration.VC will take over the management of these companies, but Sky will still retain a financial stake in them.

Regeneration.VC now counts nearly 50 companies in its global portfolio.

"We are backing these consumer companies because those categories of climate have received less funding than energy and mobility," said Rowles. "And we're seeing a really exciting trend of some of the world's biggest corporates wanting to make changes in their value chains and their supply chains."

Trump 2.0 is spurring climate funds to invest in Europe

Along with the expansion of its portfolio, Regeneration.VC is adding a new partner to its European office. Narina Mnatsakanian, who was UBS' former head of impact investment, will also be the fund's new chief impact officer. In 2023, Regeneration's founding partner, Michael Smith, moved to the Netherlands.

Rowles also joined the firm in November 2024 in a bid to double down on its presence in the UK, and raise capital from limited partners for its second $150 million fund.

Europe is fast becoming an attractive hotspot for climate VCs amid a change in the US administration. Trump has been vocal in his support for oil and gas initiatives and has threatened to push back on the Inflation Reduction Act, which facilitated investment in clean energy. A potential shift in climate policy has encouraged some climate tech investors to look to Europe for more investment opportunities, Rowles said.

In the past year, US-based climate tech funds such as Voyager Ventures and Energy Impact Partners have also grown their European footprint with new appointments on the continent.

"The concept of sustainable and impact investment will have better tailwinds in Europe for the next few years," Rowles told BI. "Can sustainable investing survive Trump 2.0? For us, it's strategic to have a European presence, even though the DNA of the firm is the US. We have to be aligned with EU systems."

Regeneration.VC will still be investing in the US because of its "great pools of capital and risk appetite," — but Rowles emphasized that being aligned with the European Union on climate was a huge priority for the fund.

"We will see more climate tech funds with more of a presence in Europe," he said. "The reality is, you need to be in both markets to do the type of work we're doing. But in Europe, there's better regulation and better mechanisms to avoid greenwashing," he added.

• Mistral plans to IPO instead of a sale, its CEO said in an interview at the World Economic Forum.
• Founded in 2023, Mistral develops open-source large language models.
• Mistral, one of Europe's top AI startups, has raised over €1 billion since its launch.

French AI startup Mistral, often billed as Europe's answer to OpenAI, plans to take the initial public offering route instead of being acquired, its cofounder and CEO Arthur Mensch said at the World Economic Forum on Tuesday.

Launched in 2023 by former DeepMind and Meta researchers, Mistral is building open-source large language models, known as LLMs, and has released a generative AI chatbot called "Le Chat." It claims its models can run much faster than its competitors.

Amid industry chatter that consolidation could be on the cards for LLM startups this year, Mensch said in an interview with Bloomberg TV in Davos, Switzerland, that Mistral "is not for sale" and that his company would instead look to IPO at some point.

Mistral is one of Europe's most valuable AI startups and the continent's only contender that has created AI models to rival the likes of OpenAI and Anthropic.

It has raised consecutive mega-rounds totaling over €1 billion since its inception — last raking in €600 million (around $620 million) in a round led by General Catalyst at a $6.2 billion valuation in June 2024.

Mensch added that Mistral would be opening an office in Singapore while doubling down on growth in the US and Europe.

"We are the only European company that is providing what we provide," Mensch said, adding that being based in Europe brought a "lot of strengths," such as a strong talent pool. "We have a tendency to play referee in a trillion-dollar race, but we have the strength — we just need to have the will to create in Europe."

Mistral recently signed a contract with a NASDAQ-listed blue-chip company and also inked a multi-million euro deal with French media company Agence France-Presse to incorporate its articles into its chatbot.

OpenAI has also signed similar content licensing deals with Axel Springer, News Corp, and the Financial Times.

Mistral started ramping up its hiring spree in Palo Alto at the end of last year in a bid to establish a transatlantic presence and attract more AI talent.

• Synthesia has secured $180 million in Series D funding led by NEA, valuing it at $2.1 billion.
• The London-based startup uses AI to create videos and avatars for businesses.
• Check out the 18-slide pitch deck the startup used to secure fresh funding.

Generative AI startup Synthesia has secured $180 million in a Series D funding round led by venture firm NEA.

The London-based startup, launched in 2017, creates AI-generated videos and avatars for businesses.

The round values the startup at $2.1 billion — making it the UK's largest generative AI startup by valuation, according to data platform Dealroom.

"Over the last few years, we think we've built the market-leading AI video platform for enterprise," cofounder and CEO Victor Riparbelli told Business Insider in an interview.

Synthesia offers its customers a suite of products, including tools for creating custom AI avatars, an AI screen recorder that can help edit videos on the startup's platform, and an AI video assistant that can turn web pages into videos.

Companies like Zoom and UiPath use Synthesia's platform to develop everything from customer service videos to corporate training clips.

Riparbelli said the company's global client base distinguishes it from other European AI companies. "Over half our revenue comes from the US, so we've been quite successful at building not just a regional player but a global player out of the UK," he said.

Synthesia has benefited from the heightened investor appetite for growth-stage AI startups. The Series D was led by NEA, with participation from existing investors, including Google Ventures, and new backers, including WiL (World Innovation Lab), Atlassian Ventures, and PSP Growth. This brings its total funding to $330 million.

With the cash injection, Synthesia will continue to hone its existing product suite and build out its video publishing platform. "We're also building some of our AI agents that help you in the video production process," Riparbelli added.

The startup also plans to expand its geographic footprint in Japan, North America, and Australia in the coming year.

Check out the 18-slide pitch deck it used to secure the fresh funds.

• Xocean's uncrewed ships offer low-carbon ocean data for offshore wind and hydrography.
• The Irish startup has secured $119 million to expand its services.
• Check out the 10-slide pitch deck it used to secure the funding.

Xocean, which offers data on the ocean to offshore wind and hydrography operations, has secured $119 million to expand.

The Irish startup, launched in 2017, has developed remotely controlled uncrewed surface vessels (USVs) with built-in sensors that can capture geophysical ocean data.

Users monitor and control the USVs through the startup's CyberDeck cloud platform, which also analyzes the quality of the data being collected. This data can give insights into seafloor topography and the sediments that make up the ocean floor.

Xocean's clients include BP, Shell, and SSE Renewables.

"We are providing this service for many of the world's largest energy companies, supporting the development of clean, renewable energy globally," founder and CEO James Ives said in a statement announcing the $119 million investment.

The transition to cleaner energy has been a central topic of discussion as companies race to find renewable sources of energy — such as wind and nuclear — to power the AI boom.

The startup partnered with climate investment firm S2G Ventures to structure the $119 million round, which was backed by S2G, Climate Investment, and Morgan Stanley's 1GT climate fund, among others.

The company said it would use the capital injection to expand its geographical footprint and accelerate product innovation.

Check out the 10-slide pitch deck used to secure the fresh funding.

• Big Tech is investing in nuclear power to meet AI data center energy demands.
• Nuclear is seen as a clean energy source, but investors are skeptical about scalability and returns.
• While VC interest in nuclear startups is growing, startups face key bottlenecks.

The generative AI boom has made nuclear power a major new obsession for Big Tech. Some industry watchers aren't fully convinced that it should be — or that nuclear startups will be able to capitalize on the hype.

This year, companies at the forefront of AI development have been in a frenzy over nuclear power as they've searched for clean sources of electricity to run the energy-hungry data centers being built to serve their prized AI models.

Microsoft made a stunning move in September as it struck a 20-year power purchase agreement with Constellation Energy to awaken one of two dormant nuclear plants on Three Mile Island — the site of one of the most high-profile nuclear accidents in US history.

In October, Amazon took a stake in X-energy, a developer of small modular reactors (SMRs) that promise greater efficiency than large nuclear reactors. That same month, Google announced a clean energy agreement with Kairos Power, a company developing SMRs.

These deals have emerged at speed for a simple reason. An arms race in the tech sector between companies vying for control of the most powerful AI models is set to drive data center power demand through the roof, with Goldman Sachs estimating a 160% jump by 2030.

However, while Big Tech's ambitions to build the world's most potent AI models have invigorated their interest in nuclear power, investors, energy experts, and analysts are feeling split about whether it will help startups scale at pace and deliver fruitful returns.

Why nuclear might not be a quick-fix solution

One issue that skeptics point to is that nuclear reactors won't come online quickly enough or with the scale needed to meet the demands of energy-hungry data centers.

Jill McArdle, a campaigner at European nonprofit Beyond Fossil Fuels, told Business Insider that nuclear power is "completely off topic" as a current solution for powering data centers, particularly if tech companies are serious about the looming deadlines they've set to meet emissions targets.

Google aims to achieve net-zero emissions across all of its operations by 2030. Microsoft, meanwhile, has committed to being carbon-negative by 2030. "What we are talking about, especially now, is the next five years of how are we going to power this massive boom in data centers," McArdle said.

She added that more compact SMRs adopted by Big Tech also remain largely untested. Google's corporate agreement with Kairos Power, for instance, is expected to see the startup's first SMR come online by 2030, with others being added through to 2035.

One concern around large nuclear solutions is expense, with the likes of Microsoft's Three Mile Island deal unlikely to be replicated elsewhere. As McArdle put it: "Traditional nuclear just isn't going to be coming online at the scale and in the budget that we need to get it done."

Venture capitalists have echoed this concern.

"The length of the investment is not compatible with private equity funds — maybe it's one for evergreen funds," said Guillaume Sarlat, partner at France-headquartered VC firm Axeleo Capital, which has deliberately excluded nuclear from its investment policy. "The other problem is, what are the economic conditions going to be when nuclear startups are ready to sell their product? What is going to be the cost of the electricity that they're going to produce in 20 years?"

He speculated that funds backing nuclear could aim for an internal rate of return of 15%, but the two main parameters that determine this would be productivity gains and the competitiveness of the nuclear solution. These factors could be affected by the price of gas and photovoltaic materials, making it a risky bet, he said.

Startups face key bottlenecks

On the technical side, nuclear startups will have to work hard to differentiate against existing fission technologies and "persuade investors that that marginal improvement is worth waiting another 10 years," said Matthew Blain, principal at climate tech fund Voyager Ventures.

While Blain noted an aligned "excitement" for nuclear fusion technology, he added that these startups would first need to demonstrate a believable pathway down the cost curve. "Your first dollars per megawatt of your first fusion plant will be astronomically expensive, and that will be competing on a 20 to 30-year timeframe with the cost of energy and battery storage," he told BI.

It's part of the reason investment in nuclear energy startups has fluctuated over the past five years. The industry had a banner year in 2021, with startups raking in $3.57 billion in VC funding, per PitchBook data. Figures subsequently dipped in 2022 and 2023, with VCs pouring $2.67 billion and $1.17 billion into startups, respectively.

"Nuclear energy requires a centralized infrastructure that is harder to scale incrementally," said Nicolas Heuzé, cofounder and CEO of osmotic energy startup Sweetch Energy. "And investors and governments often favor proven solutions, even though they are not perfect, over novel ones associated with emerging technology."

The case for being bullish on nuclear

Despite the concerns, certain quarters of the tech sector remain convinced that nuclear is the way forward to support the AI data center boom.

A16z, the venture capital firm led by Marc Andreessen and Ben Horowitz, named "the resurgence of nuclear" as one of its big ideas for driving its "American dynamism" investment theme in 2025.

"A perfect storm of regulatory reform, public enthusiasm, capital infusions, and insatiable energy needs — particularly from AI data centers — will accelerate orders for new reactors for the first time in decades," is how David Ulevitch, a general partner at A16z, put it.

A few things still need to be figured out. Blain notes that VCs will need to see if there is profit to be made on a technology that may offer "more of an infrastructure return" typically made through debt investments than the kind of outsized return a VC typically seeks from a bet on a software business. Nuclear startups may also opt to "take the trajectory of companies like SpaceX by staying private for a long period of time," he said.

That said, it's clear that money is flowing into the industry again, as VCs have deployed $2.62 billion into nuclear startups this year. Notable raises included X Energy's $500 million round and the $151 million raised by Paris-headquartered Newcleo, which is building SMRs using repurposed radioactive waste.

Newcleo's COO, Elisabeth Rizzotti, told BI that a Big Tech-fueled boom in demand for clean energy had made it an "attractive" option for investors. She added that the startup was potentially eyeing an IPO once it met two key milestones: building its first prototype in 2026 and getting pre-authorization to build its first reactor in France by early 2027.

Companies trying to sell the world on nuclear power will have to accept a hard reality, however: the clock's ticking on their opportunity to prove their solutions can meet the extraordinary energy demands of the AI industry. The data centers will keep on coming.

• Paris-headquartered Newcleo is developing small modular reactors powered by radioactive waste.
• The startup, Europe's only nuclear unicorn, is riding a wave of Big Tech interest in nuclear energy.
• Elisabeth Rizzotti, its COO, said the startup will go public after hitting two key milestones.

A three-year-old startup is banking on molten lead and radioactive waste to fuel a nuclear energy renaissance driven by the tech world's demand for abundant and clean power.

Paris-headquartered Newcleo is building small modular reactors, a type of scaled-down nuclear fission reactor assembled in a factory that can be deployed locally at operating plants. It is aiming to capitalize on the frenzy to power energy-intensive AI data centers.

The startup's unique selling point is that its SMRs use molten lead instead of water as a coolant to transfer heat from the reactor core to a power-generating system.

Notably, the startup's SMRs use spent plutonium and uranium radioactive waste as fuel by separating these two elements from the unwanted fission products.

"We're using and reducing what is considered to be a liability and a strong concern for all the governments who have so many deposits of nuclear waste," Elisabeth Rizzotti, the cofounder and chief operating officer of Newcleo, told Business Insider in an interview. "So we also promised to reduce them. And this is a story that really excites people."

Big Tech's push into nuclear has galvanized the startup's efforts. Tech giants such as Google and Amazon have been pivoting to nuclear energy sources to power data centers, which in turn power AI applications. In September, Microsoft announced a deal to help restart Three Mile Island, a nuclear plant infamous for housing one of the most significant nuclear accidents in US history.

Localized nuclear power

Newcleo is one of a handful of companies developing SMRs. Competitors include NuScale, X-Energy, and Rolls-Royce. One appeal of SMRs is that they can be readily transported to industrial sites.

"One of our most relevant business models is supporting the industrial sector, in providing them energy locally," Rizzotti said. "The small models are easily transportable because they're flexible, and you can build them in the factory, so it's very feasible — so there is strong interest from the AI sector, but also wider commercial partners."

She told BI that Newcleo has been approached by other industrial sectors, such as ceramic metal producers, in addition to those in the AI ecosystem.

Securing industrial partnerships is key to Newcleo's growth strategy and fundraising efforts. The startup aims to develop its first prototype by 2026 and operate its first reactor by 2031. Right now, Newcleo is lossmaking, but earlier this year, it hit around $55 million in revenue through its three subsidiaries.

Growth ambitions

In September, Newcleo moved its headquarters from the UK to Paris, which Rizzotti said was to be closer to the French government and support its licensing applications. The move helped Newcleo get selected to be part of the European Industrial Alliance on Small Modular Reactors, an initiative to accelerate the deployment of SMRs.

Despite public and private sector interest, nuclear energy is still seen by some as a risky enterprise due to safety concerns and the hefty upfront cost of building power plants.

Rizzotti acknowledged that Newcleo's investors had a "high risk appetite" — but says Newcleo's place in the circular economy makes it an attractive investment option.

In March 2023, Newcleo announced a bold ambition to raise €1 billion, about $1.04 billion. It's now over halfway to meeting that target, having raised over $560 million in funding from the likes of Pi Campus, Tosto Group, and Viaro Energy since its launch in 2021.

Its future capital-raising strategy will also partially depend on public funds. "At a European level, energy independence is something that we need to consider," Rizzotti said. "It is very important to have a partnership between the public and private, because public investment will reassure private investors."

The startup also sees an IPO as a viable exit option in the future. "Sooner or later, we will start the project of an IPO, so the perspective of being public allows people and investors to have shares of a company that is recognized in the market," Rizzotti told BI. "It is a tangible asset."

While Rizzotti did not have a set timeline for the company going public, she added that it would likely be after the startup achieved a second key milestone after its first prototype in 2026: pre-authorization to build its first reactor in France by early 2027.

"Once we obtain these two milestones, in our opinion, we could be ready to go public," she told BI. "Otherwise, we will continue to raise money through more natural channels like private equity or venture capital."

• Microsoft is diversifying the AI models for 365 Copilot to reduce reliance on OpenAI, per Reuters.
• The move aims to cut costs and improve speed for Microsoft's enterprise clients.
• Big Tech firms have invested heavily in AI startups to develop advanced models.

Microsoft is diversifying the artificial intelligence models it uses to power its flagship AI assistant, 365 Copilot, in a bid to reduce its dependence on OpenAI, Reuters first reported.

The Big Tech giant is moving toward adding internal and third-party AI models to help run its 365 Copilot to cut costs and address concerns about speed for its enterprise clients, per the report.

As one of OpenAI's main backers and corporate partners, Microsoft will continue working with the AI startup to develop frontier models.

"We incorporate various models from OpenAI and Microsoft depending on the product and experience," Microsoft said in a statement to Reuters.

Microsoft can customize OpenAI's model, per its original licensing agreement with the company. While Microsoft is currently training its own model, Phi-4, the software juggernaut is looking at modifying other third-party models to reduce the cost of running 365 Copilot.

Microsoft's lackluster debut of Copilot raised concerns about the software giant's ability to deliver on its AI ambitions, BI previously reported. Some customers appear to be dissatisfied with the product, spurring complaints that it is ineffective, expensive, and not secure.

In the race to develop powerful frontier models, Big Tech giants have scrambled to bulk up their arsenal of AI investments — and pumped billions of dollars into startups to help them achieve this goal.

Amazon has invested $8 billion into AI juggernaut Anthropic and used the startup's technology to power its digital assistant. This year, Google signed a deal with Character.AI, a startup that develops anthropomorphic chatbots, which allowed it to hire its founder and license its technology — a deal described as an "acquihire."

Microsoft and OpenAI did not immediately respond to a request for comment, sent outside standard working hours.