• Meta has begun to cut thousands of jobs to focus on AI investment and efficiency.
• Mark Zuckerberg is targeting low performers, part of a broader industry move toward leaner operations.
• Some employees told Business Insider they're anxious about the changes.

Meta has begun to cut thousands of jobs as the social media giant takes a tougher stance on underperforming employees and readies its finances for another year of heavy AI investment. Affected employees in Europe, Asia, and the US have started to be notified, per an internal post viewed by Business Insider.

The company has said it will eliminate roughly 5% of its workforce, which could mean almost 4,000 employees lose their jobs.

CEO Mark Zuckerberg told staff in January he would "raise the bar" and move quickly to remove low performers, according to an internal memo seen by BI.

This is part of a broader push by Big Tech companies to make themselves leaner after a hiring spree during the pandemic. Microsoft, Amazon, Salesforce, and others are collectively eliminating thousands of employees.

Zuckerberg has been at the forefront of this, announcing a "year of efficiency" in 2023 that has continued through last year and into 2025. Wall Street has rewarded Meta for this new focus, sending the company's shares soaring since the start of 2023 — a run that's added more than $1 trillion to Meta's market valuation. 

While Meta remained profitable through recent periods of heavy hiring and big spending, the company is now racing to keep up with rivals in the generative-AI race. This requires billions of dollars in infrastructure and related investment. That's likely putting pressure on Zuckerberg to seek cost savings elsewhere. 

A Meta spokesperson declined to comment.

Impact on some employees

Meta is offering impacted workers in the US a severance package that includes 16 weeks of pay and an additional two weeks for every year of service, according to two people familiar with the matter.

For some Meta employees, the efficiency drive is causing anxiety. These staffers asked not to be identified discussing sensitive topics.

"Mark is creating fear," a Meta employee told BI. "He's creating a culture where you have to be loyal to him or else."

Another employee said that working at Meta right now "feels like living in a George Orwell novel."

Even colleagues who have performed well "have been disappearing all year, and when you ask about it, you're just told, 'They're no longer with the company,'" this person said. "Self-censorship is rampant. At a company supposedly dedicated to connecting people, the human side of our work is disappearing, and everyone is acting more robotic."

Another Meta employee said reductions shouldn't be branded as performance-based cuts because this could damage people's reputations as they seek other opportunities.

"Now people have to go back out into the job market with a label that is incredibly unfair," this person added.

They expressed concern that good employees would be cut just to meet quotas and that this could have a negative impact on morale.

"What's the incentive to help a new hire ramp up if they're just going to stack rank us and probably do this all again next year?" this person added.

How Meta's latest job cuts may work

The job cuts are designed to target employees who receive "met some" or "did not meet" ratings, the bottom two categories in Meta's assessment system, in their performance reviews.

Internal guidance obtained by BI last month says managers must identify 12% to 15% of employees eligible for these ratings. Meta aims to reach 10% "nonregrettable attrition" by combining these cuts with previous departures. For example, if a team had 5% attrition in 2024, managers would need to identify another 7% to 10% of their employees for the bottom ratings to meet the target.

One Meta employee told BI that forcing managers to place team members into bottom categories for job cuts had spread anxiety through the management ranks as well as the rank and file.

On Friday, employees received a memo from Janelle Gale, Meta's vice president of human resources, detailing how the process should work. The memo, which was obtained by BI, said affected employees would be notified through their work and personal email addresses and lose access to company systems within an hour of being informed. They'll receive information on their severance packages in the same email, it added.

The notifications will be staggered across time zones, with employees in the Asia Pacific region being notified first, followed by those in Europe, the Middle East, and Africa, and then, finally, North and Latin America, the memo said.

Employees in European countries such as Germany, France, Italy, and the Netherlands will be exempt from this process because of local regulations and will instead follow local performance management processes, the memo said. Meta intends to backfill these roles, it added, but plans and timelines "may vary."

How Meta is reorganizing itself

Amid the cuts, the social media giant is also reorganizing some of its businesses and divisions.

The company is merging its Facebook and Messenger teams under Facebook's chief, Tom Alison, while Messenger's head, Loredana Crisan, is set to move to the generative-AI group, The Information said.

Meta's Reality Labs division, which has lost nearly $60 billion since 2020, is being more tightly integrated with Meta's main business, reversing some of Zuckerberg's 2021 reorganization. In an internal memo obtained by BI, Reality Labs' chief technology officer, Andrew Bosworth, said Reality Labs had "become a positive driver for Meta's overall brand."

• Meta has enacted rules that bar employees from discussing politics, health, and weapons at work.
• These Community Engagement Expectations restrict mocking topics within a protected category, like race.
• Some employees formed CEE Watch to monitor post removals.

Some Meta employees are questioning the company's removal of their posts and comments from its internal forum.

Employees created a group on its internal Workplace forum to share their experiences in posts that have been seen by Business Insider.

In 2022, Meta rolled out internal rules barring employees from discussing contentious topics such as politics, health, and weapons. The guidelines, which Meta calls Community Engagement Expectations, also ban comments and posts that are seen as mocking topics related to a protected category, such as race, gender, or religion.

Some Meta employees have accused the company of using the CEE system to censor valid discussions. A number of staffers recently created an internal employee resource group called CEE Watch so that Meta staff could flag when a post has been removed, according to internal documents seen by BI.

"CEE language is intentionally vague and we cannot know how it's being enforced without openly sharing our violations with each other," a welcome post on the CEE Watch page says.

CEE Watch had over 800 members at the time of publishing, a small fraction of Meta's 72,000-strong workforce.

A Meta spokesperson said the company didn't remove internal employee comments just because it doesn't agree with or like them. They added that many critical comments remained up on Meta's internal communications boards.

One employee wrote that multiple posts regarding Palestinians had been removed over the past year. The person questioned how employees could engage in discussions of "what are acceptable forms of identity to discuss at work."

Another person commented, saying that it was an "open secret" in the "Muslim@" employee resource group that "relatively" harmless posts had been removed.

A different person added that their post about grieving slain family members was removed.

While free speech is a fundamental constitutional right in public spaces, it does not extend to private corporations. As private entities, companies have the legal authority to establish their own policies regarding what employees can or cannot say in the workplace.

But US employees are allowed to discuss wrongdoing by their companies and other workplace-related issues, like safety, harassment, and accommodations for people with disabilities.

The CEE guidelines were last updated in October, according to a copy of them viewed by BI.

They say it's "not okay" for employees to share content that has the "potential to trigger disruptive comments" around topics including: "political movements or causes relating to states, nations, or people (e.g., opinions on forms of governments, political systems, or economic systems; sharing national flags in ways that imply opinions on political movements or causes or slogans like 'Free Puerto Rico'; 'Liberate Hong Kong'; 'Make America Great Again'; 'Build Back Better'; 'Free Palestine'; or 'Ukraine today, Taiwan tomorrow!')."

In a separate post, titled "communications around LGBTQ healthcare and health plans," one employee said they asked Meta's employment law group if LGBTQ+ employees and their children would continue to have access to gender-affirming care after Meta's public benefits page removed mention of it.

The person said they "heard back that there were no plans to update the plan, but that they would comply with applicable laws."

Another person said a drag performance group, which did an act for one of the company's pride employee resource groups, was investigated by the company and later banned because it claimed the group violated the CEE guidelines. This person said the performance was livestreamed across the ERG and was met with positive reviews.

"This feels like a disproportional punishment for this performer and impacts their livelihood," the person wrote. "I'm currently escalating this case now, but I thought it would be helpful to share that CEE also applies to guests of Meta (apparently even retroactively too)."

In a comment under the post, the original poster claimed that during his in-person conversation with the internal community relations team, "they said that 'those kinds of people are high risk for violating content' (referring to drag performers)." The poster added: "I left the conversation extremely upset."

This person shared a message from the internal community relations team that said the drag group's performance in 2022 violated the CEE and other company policies, "including multiple remarks degrading about various protected categories, and multiple instances of sexual content shared."

Meta employees are increasingly vocal about the company's content moderation on its internal forums, with some directly challenging what they view as censorship of workplace discussions.

"This is a free speech issue," one employee wrote after one of their posts was removed by Meta's internal moderators. It linked to a news article about Donald Trump saying that Meta's CEO was "probably" changing the company's direction in response to Trump's previous threats to jail him.

The employee questioned how Meta, a company meant to be "hardened against threats," could restrict internal discourse about its own leadership.

A six-year employee in Meta's civic integrity team described deteriorating trust between leadership and staff.

"When you tell people they should resign if they don't agree with your decisions, this belies a lack of trust in your people," they wrote.

The CEE guidelines say "consequences for violating this policy vary depending on the severity of the violation and other context such as a person's prior conduct." They list disciplinary action, including termination of employment, as one example.

One employee in Reality Labs, Meta's virtual reality division, said it was ironic to have "debate openly" as a core principle as posts were removed.

"Attempts to even have the most polite discussion that acknowledges how people are responding to policy changes" were deleted, they wrote. They also expressed concern about whether there would be retaliation in Meta's coming layoffs for discussing workplace conditions.

More employees wrote that there was a climate of fear around posting, with content sometimes being removed within minutes without explanation.

One worker said colleagues deleted comments and self-censored "because they were afraid they might be targeted by CEE and management for even reacting positively to a post."

"I have very mixed feelings about Meta at this point," the civic integrity employee wrote, a sentiment that was shared across multiple posts.

"The greatest value of this company is not its products or its technology but its people. They make all of it possible," this employee said. "The leadership of this company maybe understood that at some point, but they seem to have forgotten it."

Are you a Meta employee? Got insight to share? Contact the reporter Jyoti Mann via email at [email protected] or via Signal at jyotimann.11. Reach out from a nonwork device.

If you're a current or former Meta employee, contact Pranav from a nonwork device securely on Signal at +1-408-905-9124 or email him at [email protected].

Got a tip? You can reach Hugh using the secure messaging app Signal (+1 628-228-1836) or secure email ([email protected]).

• Meta kicked off 2025 with big changes following a big year for the company.
• In the first weeks of January, it overhauled its content moderation system, nixed DEI programs, and announced layoffs.
• The "rather frenzied start" to the new year sets up an apparent "year of intensity," one analyst told BI.

Coming off of a banner year in 2024, Meta is hitting the ground running in 2025.

With a new president in the White House and the AI arms race in full swing amid DeepSeek mania, CEO Mark Zuckerberg has kept busy.

Referring to Meta's January as "a rather frenzied start," Forrester VP and research director Mike Proulx said that the company's past "year of efficiency" was being "trumped in 2025 with an apparent year of 'intensity.'"

The company made several big announcements in the first week of the new year ahead of Trump's inauguration, and in audio from Meta's all-hands meeting this week obtained by Business Insider, Zuckerberg told employees to "buckle up."

"Everyone always says that every year is a big year, right?" Zuckerberg said.

"When I look at the kind of long-term trajectory for the stuff that we're doing this year, I think by the end of this year we're going to have a much clearer sense of the trajectory of a lot of the long-term things that we're doing, whether that's AI or glasses, or a number of areas around the future of social media, a number of key partnerships that we're working on," the Meta CEO said.

"This year feels a little more like a sprint to me," he added.

New year, new policies that will impact your News Feed

If 2025 represents a sprint for Meta, the first leg of the race saw some of its biggest changes in years.

The most impactful for Meta's billions of users was announced on January 7, when Zuckerberg messaged that Meta's content moderation policy would be overhauled.

In a controversial decision, Meta said that it would sunset its third-party fact-checking in favor of community notes, similar to that of Elon Musk's X, formerly known as Twitter.

Zuckerberg said Wednesday he thinks a community notes model "is just going to be more effective."

The company also loosened its hate speech policies and announced that it would return to promoting political discussions in its apps.

The policy changes may not move the needle much on the ad dollars spent on Meta's platforms as its size alone makes it a "must-have" for advertisers, Truist Securities analyst Youssef Squali told BI.

"Large advertisers will tell you that. Small advertisers will tell you that," the analyst said. "So there's scarcity value for advertisers and their ability to access some of these platforms that actually work for them."

Meta CFO Susan Li said in Wednesday's earnings call that the company hasn't observed a noticeable difference in ad spending since the announcement a few weeks ago. When reached for comment, Meta referred BI to Zuckerberg's video and its blog post on the content moderation changes.

Collectively, Meta's policy overhaul means big changes are coming to your Instagram, Threads, and Facebook news feeds.

Meta maneuvers closer to Trump

The content moderation policy overhaul is one of a slew of recent changes at Meta widely viewed as a way to appease President Trump, who frequently criticized and even threatened the company and its chief executive in the past.

Trump told reporters the content moderation changes were "probably" in response to threats he's made to Meta.

Following a dinner with the president at Mar-a-Lago and Meta's $1 million donation to his inauguration fund, Zuckerberg was one of a handful of tech execs seated close to Trump during his swearing-in ceremony on January 20.

"Having Trump coming back into office — and we all know Trump's position with regards to Meta and Zuck, in particular — I feel there was a lot of pressure for him to revisit his stance," Truist Securities' Squali told BI of the content moderation changes.

Meta's chief marketing officer, Alex Schultz, previously told BI that Trump's election win, coupled with shifting "vibes in America," were factors in the content moderation overhaul.

Other big changes include the company rolling back its programs for DEI. Diversity programs have been a frequent target for Trump and other conservative activist groups.

In Wednesday's earnings call, Zuckerberg said 2025 would be "a big year for redefining our relationship with governments." He added that he's "optimistic" about the Trump administration, saying it "prioritizes American technology winning" and "will defend our values and interests abroad."

Meta also made some personnel changes that could help it better navigate the next four years.

Ahead of the administration change, Meta named Joel Kaplan its new chief global affairs officer. Kaplan, a former adviser to George W. Bush and a longtime Republican lobbyist for Meta, replaced Nick Clegg, former leader of the UK's Liberal Democrats, in the position.

Meta also added several new names to its board of directors, specifically tech investor Charlie Songhurst, Exor CEO John Elkann, and UFC chief Dana White. Zuckerberg, an MMA hobbyist, has been photographed with White, a longtime friend of Trump, multiple times in recent years.

And in a recent move to end a legal battle involving Trump, Meta also agreed to pay $25 million to settle a lawsuit the president brought against it and Zuckerberg after Facebook suspended his account in 2021 following the January 6 Capitol riots.

Plans for layoffs and doubling down on AI

Meta's work to reshape its workforce, which began in late 2022, shows no signs of slowing down as it races to keep up with AI pioneer OpenAI and newcomers like Chinese AI startup DeepSeek.

In a memo obtained earlier this month by BI, Meta announced layoffs coming in February for 5% of its workforce, targeting its "lowest performers."

However, discussing its quarterly earnings on Wednesday, Meta said it plans to hire in "the priority areas of infrastructure, monetization, Reality Labs, generative artificial intelligence (AI), as well as regulation and compliance."

The company expects to spend $60 to $65 billion in capital expenditures this year as it barrels full-steam ahead on AI, a significant increase from $39.23 billion in 2024, which already raised eyebrows among investors and analysts.

Zuckerberg said Wednesday he thinks 2025 will be the year an AI assistant hits 1 billion users, and naturally he wants that of Meta AI. He also teased upcoming news about its Llama 4 AI model and hinted at getting back to "some OG Facebook" this year.

Responding to Wall Street analysts' questions about DeepSeek, Zuckerberg said it was important for a US firm to set the standard on open-source AI "for our own national advantage."

"I think that, if anything, some of the recent news has only strengthened our conviction that this is the right thing for us to be focused on," Zuckerberg said.

In audio from Meta's all-hands meeting on Thursday, Meta CFO Susan Li said the company is "excited for the roadmap" this year and for AI initiatives "driving further momentum" that would "fuel 2025 growth."

Do you work at Meta? Contact the reporters from a non-work email and device at [email protected] and [email protected].

• Mark Zuckerberg addressed several recent policy changes in a company all-hands meeting on Thursday.
• Zuckerberg admitted that the company was slow to respond to TikTok's rapid growth.
• The CEO also told employees to "buckle up" for an "intense" year ahead.

Meta CEO Mark Zuckerberg acknowledged the company was slow to respond to TikTok's meteoric rise because executives didn't view it as truly social, offering a rare window into how the tech giant missed one of social media's biggest shifts in recent years.

"When I look back on TikTok, I think part of the reason why we were slow to it is because we didn't think TikTok was social," Zuckerberg said in a recording of an all-hands meeting obtained by Business Insider. "We looked at it and we thought, 'Oh, this is like, a little more like YouTube.'"

The admission came in response to an employee's question about whether Meta's current focus on artificial intelligence might cause the company to miss the next major social media trend, as it did with TikTok.

Meta did not immediately respond to a request for comment from Business Insider.

Zuckerberg explained at the meeting that Meta's traditional view of social interaction — centered around friends posting content and commenting — caused the company to initially misread TikTok's appeal. The company failed to recognize how users were sharing TikTok content through private messages, which has become a crucial form of social interaction across Meta's platforms.

"Because we were too dismissive up front, it wasn't just about people commenting in the feed. It was about people seeing stuff in their feed and then sharing it into message threads," Zuckerberg said, referring to the company's instant messaging platforms WhatsApp, Messenger, and direct messaging in Instagram, where "the majority of social interaction is happening."

Zuckerberg also addressed TikTok's uncertain future in the US. President Trump signed an executive order on January 20 that gave TikTok 75 more days to operate in the US, as owner ByteDance will either have to divest from TikTok or it will be banned in the US.

"We don't have control of what's going to happen to Tiktok," Zuckerberg said. "We have a lot of competitors, but they're an important one. So, who's gonna own Tiktok at the end of the year? What's gonna happen? I mean, that's a pretty big deal, something that's a card that we get to turn over."

Looking ahead, Zuckerberg emphasized that the company needs to avoid taking "too narrow of a view" of social interaction as it navigates the emergence of AI. He outlined a vision for AI-powered features in Facebook and Instagram feeds, including interactive AI agents that users can converse with and more immersive content experiences.

"I think the next trend here is there're going to be things that either AI can produce, that we can just put in there... I think this year we're gonna have stuff where it's like, okay, you have an AI agent, and you can just start talking to it," Zuckerberg said.

The Meta chief also pushed back against concerns that the company's AI investments might detract from its core social media business, noting that as a large company, Meta needs to be able to "walk and chew gum at the same time."

"If we can't build Facebook and [the] next platform at the same time, then, like, eventually game over," he said.

Do you work at Meta? Contact this reporter from a nonwork email and device at [email protected] or [email protected]. You can also reach him securely via Signal at +1408-905-9124. Your identity will be protected.

• Meta CEO Mark Zuckerberg addressed Chinese AI startup DeepSeek's latest model in an all-hands.
• The Meta chief praised its "novel infrastructure optimization advances" in a recording seen by BI.
• Zuckerberg also said it would "benefit" Meta as it can implement some of DeepSeek's methods.

Though DeepSeek may have caused panic among tech investors this week, Meta executives told staff on Thursday that the Chinese startup's breakthrough could ultimately benefit Meta's AI ambitions in the long run.

In a Q&A during a Thursday company all-hands, Meta CEO Mark Zuckerberg was asked for his thoughts on DeepSeek and how Meta should respond or pivot its infrastructure spend to set itself up for success, according to a recording reviewed by Business Insider.

Zuckerberg responded that he doesn't think it changes the company's infrastructure spend.

"DeepSeek had a few pretty novel infrastructure optimization advances, which, fortunately, they published them, so we can not only observe what they did, but we can read about it and implement it, so that'll benefit us," Zuckerberg said.

Zuckerberg also said that it's "always interesting when there's someone who does something better than you" and that it motivates him to "make sure we're on it."

Last week, Zuckerberg announced in social media posts on Meta platforms that the company will spend $60 billion to $65 billion in additional capital investment for the year and said 2025 will be a "defining year for AI."

Nvidia, which builds AI chips, plunged in value this week amid a $1 trillion tech sell-off in the markets over fears that DeepSeek had leapfrogged US AI labs, as the Chinese startup said its R1 model rivaled OpenAI's o1 across several benchmarks and that it was built for a fraction of the cost and compute power.

DeepSeek's claims have caused investors to question Big Tech firms' spending on AI infrastructure, and shares plummetted. By Wednesday, tech stocks had partially recovered.

Meta CFO Susan Li told staff on Thursday that Meta was "seen as more insulated" from DeepSeek's market impact than other companies because it didn't rely on selling its own open-source AI model, Llama, to generate revenue.

"Does it mean that over the longer run, Meta will have to spend less on infrastructure investment or get more for their existing their existing investment?" Li said. "That's definitely sort of what I've heard so far in a few calls since the earnings call yesterday with our analysts and our investors in terms of parsing where reactions are."

Meta declined to comment.

Do you work at Meta? Contact BI reporters from a nonwork email and device at [email protected], [email protected], and [email protected]. You can also reach them via Signal at jyotimann.11, hughlangley.01, and +1408-905-9124.

• Meta CEO Mark Zuckerberg told staffers in a Thursday all-hands to be ready for an "intense year."
• Zuckerberg addressed Meta's rollback of its DEI programs and work with the US government.
• He said that AI could lead to some roles becoming redundant, as well as hiring more engineers.

Meta CEO Mark Zuckerberg told employees Thursday in a company all-hands meeting to "buckle up" for an "intense" year ahead and addressed several recent policy changes.

Zuckerberg opened the all-hands by emphasizing a sense of urgency for the year. He told staff that he expected to have a clearer sense of the company's trajectory by the end of 2025 and that AI would be top of mind. He also addressed recent policy changes related to fact-checking and programs for diversity, equity, and inclusion.

"This is a marathon, not a sprint," he said in a recording reviewed by Business Insider. "But honestly, this year feels a little more like a sprint to me."

Meta declined to comment.

Meta is betting on AI

In a wide-ranging opening monologue, Zuckerberg predicted that 2025 would be the year a "highly intelligent and personalized" digital assistant reached 1 billion users.

"I think whoever gets there first is going to have a long-term, durable advantage towards building one of the most important products in history," Zuckerberg said, according to the recording.

Zuckerberg also reiterated his belief that this would be the year Meta started seeing AI agents take on work, including writing software. Asked whether this would lead to job cuts, Zuckerberg said it was "hard to know" and that while it may lead to some roles becoming redundant, it could lead to hiring more engineers who can harness artificial intelligence to be more productive.

"The nature of what engineering is in the future will be different than it is today," he said.

Meta made recent policy changes

Zuckerberg touched on several flashpoints in recent weeks from inside the company, including the announcement that it would move away from third-party fact-checkers to a community-notes system like that used by Elon Musk's X.

He told staff to wait and see how the new system would be implemented.

"I'm actually quite optimistic that this is going to end up being a better system," he said.

Zuckerberg also said this would be a year for "resetting" Meta's relationship with governments worldwide.

"After the last several years, we now have an opportunity to have a productive partnership with the United States government, and we're going to take that," Zuckerberg said. "I think it's the right thing to do because there are several areas, even if we don't agree on everything, where we have common cause for things that are going to make it so that we can serve our community better, and we can advance the interests of our country together, " he said, adding that Meta would do so in ways that didn't compromise its "principles or values."

Meta rolled back DEI programs

The Meta chief also addressed the company's recent changes in its stance on DEI policies and the rollback of those programs in response to a shifting legal and regulatory landscape under the Trump administration.

"The way to think about this is we're in the middle of a pretty rapidly changing policy and regulatory landscape that increasingly views any policy that might advantage any one group of people over another as something that is unlawful, and because of that, we need to adjust, or else we'll just be out of alignment with what the law is saying," Zuckerberg said.

He again said that Meta viewed diversity as a strength, adding: "Historically, we've had a handful of specific programs that were very focused on certain underrepresented groups, and I think the direction of the policy and regulatory and legal direction on a lot of the stuff is that you can't do things that advantage specific groups, even if you're trying to make up for other things."

Do you work at Meta? Contact BI reporters from a nonwork email and device at [email protected], [email protected], and [email protected].

You can also reach them via Signal at jyotimann.11, hughlangley.01, and +1-408-905-9124.

• DeepSeek, a Chinese open-source AI firm, is taking over the discussion in tech circles.
• Tech stocks, especially Nvidia, plunged Monday.
• Companies leading the AI boom could be in for a reset as DeepSeek upends the status quo.

DeepSeek, a Chinese company with AI models that compete with OpenAI's at a fraction of the cost, is generating almost as many takes as tokens.

Across Silicon Valley, executives, investors, and employees debated the implications of such efficient models. Some called into question the trillions of dollars being spent on AI infrastructure since DeepSeek says its models were trained for a relative pittance.

"This is insane!!!!" Aravind Srinivas, CEO of startup Perplexity AI, wrote in response to a post on X noting that DeepSeek models are cheaper and better than some of OpenAI's latest offerings.

The takes on DeepSeek's implications are coming fast and hot. Here are eight of the most common.

Take 1: Generative AI adoption will explode

"Jevons paradox strikes again!" Microsoft CEO Satya Nadella posted on X Monday morning. "As AI gets more efficient and accessible, we will see its use skyrocket, turning it into a commodity we just can't get enough of."

The idea that as tech improves, whether smarter, cheaper, or both, it will only bring in exponentially more demand is based on a 19th-century economic principle. In this case, the barrier to entry for companies looking to dip their toe into AI has been high. Cheaper tools could encourage more experimentation and further the technology faster.

"Similar to Llama, it lowers the barriers to adoption, enabling more businesses to accelerate AI use cases and move them into production." Umesh Padval, managing director at Thomvest Ventures told Business Insider.

That said, even if AI grows faster than ever, that doesn't necessarily mean the trillions of investments that have flooded the space will pay off.

Take 2: DeepSeek broke the prevailing wisdom about the cost of AI

"DeepSeek seems to have broken the assumption that you need a lot of capital to train cutting-edge models," Debarghya Das, an investor at Menlo Ventures told BI.

The price of DeepSeek's open-source model is competitive — 20 to 40 times cheaper to use than comparable models from OpenAI, according to Bernstein analysts.

The exact cost of building DeepSeek models is hotly debated. The research paper from DeepSeek explaining its V3 model lists a training cost of $5.6 million — a harrowingly low number for other providers of foundation models.

However, the same paper says that the "aforementioned costs include only the official training of DeepSeek-V3, excluding the costs associated with prior research and ablation experiments on architectures, algorithms, or data." So the $5 million figure is only part of the equation.

The tech ecosystem is also reacting strongly to the implication that DeepSeek's state-of-the-art model architecture will be cheaper to run.

"This breakthrough slashes computational demands, enabling lower fees — and putting pressure on industry titans like Microsoft and Google to justify their premium pricing," Kenneth Lamont, principal at Morningstar, wrote in a note on Monday.

He went on to remind investors that with early-stage technology, assuming the winners are set is folly.

"Mega-trends rarely unfold as expected, and today's dominant players might not be tomorrow's winners," Lamont wrote.

Dmitry Shevelenko, the chief business officer at Perplexity, a big consumer of compute and existing models, concurred that Big Tech players would need to rethink their numbers.

"It certainly challenges the margin structure that maybe they were selling to investors," Shevelenko told BI. "But in terms of accelerating the development of these technologies, this is a good thing." Perplexity has added DeepSeek's models to its platform.

Take 3: Considering a switch to DeepSeek

On Monday, several platforms that provide AI models for businesses— Groq and Liquid.AI to name two — added DeepSeek's models to their offerings.

On Amazon's internal Slack, one person posted a meme suggesting that developers might drop Anthropic's Claude AI model in favor of DeepSeek's offerings. The post included an image of the Claude model crossed out.

"Friendship ended with Claude. Now DeepSeek is my best friend." the person wrote, according to a screenshot of the post seen by BI, which got more than 60 emoji reactions from colleagues.

Amazon has invested billions of dollars in Anthropic. The cloud giant also provides access to Claude models via its Amazon Web Service platform. And some AWS customers are asking for DeepSeek, BI has exclusively reported.

"We are always listening to customers to bring the latest emerging and popular models to AWS," an Amazon spokesperson said, while noting that customers can access some DeepSeek-related products on AWS right now through tools such as Bedrock.

"We expect to see many more models like this — both large and small, proprietary and open-source — excel at different tasks," the Amazon spokesperson added. "This is why the majority of Amazon Bedrock customers use multiple models to meet their unique needs and why we remain focused on providing our customers with choice — so they can easily experiment and integrate the best models for their specific needs into their applications."

Switching costs for companies creating their own products on top of foundation models are relatively low, which is generating a lot of questions as to whether DeepSeek will overtake other models from Meta, Anthropic, or OpenAI in popularity with enterprises. (It's already number one in Apple's app store.)

DeepSeek, however, is owned by Chinese hedge fund High-Flyer and the same security concerns haunting TikTok may eventually apply to DeepSeek.

"While open-source models like DeepSeek present exciting opportunities, enterprises—especially in regulated industries—may hesitate to adopt Chinese-origin models due to concerns about training data transparency, privacy, and security," Padval said.

Security concerns aside, the software companies that sell APIs to businesses have been adding DeepSeek throughout Monday.

Take 4: Infrastructure players could take a hit

Infrastructure-as-a-service companies, such as Oracle, Digital Ocean, and Microsoft could be in a precarious position should more efficient AI models rule in the future.

"The sheer efficiency of DeepSeek's pre and post training framework (if true) raises the question as to whether or not global hyperscalers and governments, that have and intend to continue to invest significant capex dollars into AI infrastructure, may pause to consider the innovative methodologies that have come to light with DeepSeek's research," wrote Stifel analysts.

If the same quantity of work requires less compute, those selling only compute could suffer, Barclays analysts wrote.

"With the increased uncertainty, we could see share price pressure amongst all three," according to the analysts.

Microsoft and Digital Ocean declined to comment. Oracle did not respond to a request for comment in time for publication.

Take 5: Scaling isn't dead, it's just moved

For months, AI luminaries, including Nvidia CEO Jensen Huang have been predicting a big shift in AI from a focus on training to a focus on inference. Training is the process by which models are created while inference is the type of computing that runs AI models and related tools such as ChatGPT.

The shift in computing's total share to inference has been underway for a while, but now, change is coming from two places. First, more AI users means more inference demand. The second is that part of DeepSeek's secret sauce is how improvement takes place in the inference stage. Nvidia took a positive spin, via a spokesperson.

"DeepSeek is an excellent AI advancement and a perfect example of Test Time Scaling. DeepSeek's work illustrates how new models can be created using that technique, leveraging widely-available models and compute that is fully export control compliant," an Nvidia spokesperson told BI.

"Inference requires significant numbers of NVIDIA GPUs and high-performance networking. We now have three scaling laws: pre-training and post-training, which continue, and new test-time scaling."

Take 6: Open-source changes model building

The most under-hyped part of DeepSeek's innovations is how easy it will now be to take any AI model and turn it into a more powerful "reasoning" model, according to Jack Clark, an Anthropic cofounder, and a former OpenAI employee, wrote about DeepSeek in his newsletter Import AI on Monday.

Clark also explained that some AI companies, such as OpenAI, have been hiding all the reasoning steps that their latest AI models take. DeepSeek's models show all these intermediate "chains of thought" for anyone to see and use. This radically changes how AI models are controlled, Clark wrote.

"Some providers like OpenAI had previously chosen to obscure the chains of thought of their models, making this harder," Clark explained. "There's now an open-weight model floating around the internet which you can use to bootstrap any other sufficiently powerful base model into being an AI reasoner. AI capabilities worldwide just took a one-way ratchet forward."

Take 7: Programmers still matter

DeepSeek improved by using novel programming methods, which Samir Kumar, co-founder and general partner at VC firm Touring Capital, saw as a reminder that humans are still coding the most exciting innovations in AI.

He told BI that DeepSeek is "a good reminder of the talent and skillset of hardcore human low-level programmers."

Got a tip or an insight to share? Contact BI's senior reporter Emma Cosgrove at [email protected] or use the secure messaging app Signal: 443-333-9088.

Contact Pranav from a nonwork device securely on Signal at +1-408-905-9124 or email him at [email protected].

You can email Jyoti at [email protected] or DM her via X @jyoti_mann1

• Meta said in internal memos that it would cut 5% of workers, focusing on its lowest performers.
• Staffers asked how the cuts would work. One criticized the announcement after a "pretty rough" week.
• Some employees questioned the accuracy of internal performance ratings.

Mark Zuckerberg told Meta employees on Tuesday that the company would make "extensive" cuts to low performers, affecting about 5% of the company's workforce. For some staffers, the cuts raised questions about how performance would be measured and concerns about morale.

Employees posted their reactions on an internal message board, seen by Business Insider. Several asked how the cuts would work in practice. While some employees expressed support for raising Meta's standards, others wanted specifics about the process, particularly at a time when Meta is cutting third-party fact-checkers and rolling back DEI policies.

"This is going to be an intense year, and I want to make sure we have the best people on our teams," Zuckerberg told staffers in a companywide note on Tuesday. He said that while the company would typically remove low performers over the course of a year, it would speed up that process for this review cycle.

A separate memo sent to Meta managers, a copy of which BI reviewed, said the company planned to cut about 5% of its workforce. Meta said in its latest earning report that it employed 72,404 people globally at the end of September, meaning the cuts would be equivalent to about 3,600 roles.

Some employees said they supported the decision to raise the company's standards. One wrote, "I realize there is a need to raise the bar and I absolutely support this." Others wanted to know more about how the cuts would be decided.

"How much of this decision will be based on performance and how much will be based on area of investment," one employee asked. Another asked whether the cuts would affect "exploratory work" at Meta.

"Year of intensity?" another staff member wrote, seemingly a joking reference to Meta's "year of efficiency."

"How will HR determine the level of optimism about someone's future at the company?" another question said, appearing to refer to Zuckerberg's comment that Meta "won't manage out everyone who didn't meet expectations for the last period if we're optimistic about their future performance."

Zuckerberg said that employees in the US affected by the cuts would be told by February 10, while those in other territories could take more time. At least 664 employees reacted to Zuckerberg's post with a shocked emoji, 386 with a like, and 66 with a crying emoji.

It's been an eventful week for Meta employees. On January 7, Zuckerberg announced that the company would replace third-party fact-checkers with a community-notes system and that it planned to bring political content back to the news feed. On Friday, staffers were informed that Meta would roll back its DEI programs.

"Teams aren't exactly strong when morale is low, and this is already a pretty rough ride over the last week," one employee wrote in the internal message board on Tuesday following the announcement of cuts.

A Meta spokesperson didn't comment on the memos and the employee reactions.

Monkeys 'throwing darts'

To determine which employees it could cut, Meta will use its internal performance-review program, which scores employees in buckets such as "did not meet expectations" and "exceeded expectations."

One employee writing in the internal channel expressed skepticism that the performance-review system was an accurate way to do this. "I would say the ratings and process creates marginally better than 'monkey's throwing darts' in terms of reflecting the actual performance and impact for a large majority of the people," they wrote.

Others asked if the changes would apply to low-performing managers and how people on maternity or mental-health leave would be affected.

"How are we going to balance false positive terminations of people who have context about the products and are just having unlucky halves with the cost of ramping up new people who have an even bigger chance of not being able to perform at the desired bar?" another employee asked.

Janelle Gale, Meta's vice president of human resources, told employees in a comment that an FAQ would be posted later on Tuesday "with additional information around how this process will work."

One employee suggested LGBTQ+ employees might have "additional concerns" about their performance ratings.

"As the company builds towards broader cognitive diversity, is there any outsized weighting for LGBTQ+ metamates in these perf reviews or in who gets offered severance?" they wrote. "Just want to check whether that will be a factor."

"Absolutely not," Gale said. "This is not in any way intended to target the LGBTQ+ community or any group. Objectivity and integrity are crucial aspects of the Perf@ process and we work hard to remove bias from our systems. We do not tolerate discrimination in any way. Full stop."

Do you work at Meta? Contact BI reporters from a nonwork email and device at [email protected], [email protected], and [email protected].

You can also reach them via Signal at jyotimann.11, hughlangley.01, and +1408-905-9124.

• Mark Zuckerberg told staff Tuesday that Meta will make "more extensive performance-based cuts."
• Meta plans to cut 5% of its lowest performers, according to a separate memo sent to managers.
• Last week, Meta said it would roll back DEI initiatives and get rid of third-party fact-checkers.

Meta is planning to cut 5% of its lowest performers, Business Insider has learned.

Mark Zuckerberg has told employees he "decided to raise the bar on performance management" and act quickly to "move out low-performers," according to an internal memo seen by BI.

The Meta chief wrote in a post on Workplace, the company's internal forum, that the company will make "more extensive performance-based cuts" during this year's performance review cycle.

Two Meta employees told BI that the company kickstarted its annual performance review process last week by having employees submit their self-reviews, peer reviews, and manager reviews.

Zuckerberg said that impacted employees in the US would be notified on February 10.

In a separate memo seen by BI, Meta's director of people development growth programs, Hillary Champion, told managers that the changes would mean "exiting approximately 5% of our lowest performers."

Meta employed 72,404 people globally as of September 30, 2024, according to its most recent earnings report.

Champion added that employees terminated for performance will '"receive generous severance packages, in line with previous cuts

Meta last week said it would dismantle its DEI-focused team and scrap diversity programs in its hiring process. The company's vice president of human resources, Janelle Gale, said in the memo announcing the changes that the term DEI has "become charged" partly because it is "understood by some as a practice that suggests preferential treatment of some groups over others."

It also made changes to its content moderation policies earlier that week, including getting rid of third-party fact-checkers in favor of a community notes model similar to X.

In March 2023, Meta laid off 10,000 workers — the second wave of cuts at the company in four months — in what Zuckerberg called Meta's "year of efficiency."

Meta didn't immediately respond to a request for comment from Business Insider.

Read the full memo below:

Meta is working on building some of the most important technologies in the world - - AI, glasses as the next computing platform, and the future of social media. This is going to be an intense year, and I want to make sure we have the best people on our teams.

I've decided to raise the bar on performance management and move out low-performers faster. We typically manage out people who aren't meeting expectations over the course of a year, but now we're going to do more extensive performance-based cuts during this cycle - - with the intention of backfilling these roles in 2025. We won't manage out everyone who didn't meet expectations for the last period if we're optimistic about their future performance, and for those we do let go we'll provide generous severance in line with what we've provided with previous cuts.

We'll follow up with more guidance for managers ahead of calibrations. People who are impacted will be notified on February 10 -- or later for those outside the US.

Letting people go is never easy. But I'm confident this will strengthen our teams and help us build leading technology to enable the future of human connection.