• Pharmaceutical giants such as J&J, Merck, and Eli Lilly are embracing AI and prioritizing upskilling.
• They hope that training thousands of employees on generative AI will boost productivity in drug development.
• This article is part of "AI in Action," a series exploring how companies are implementing AI innovations.

Johnson & Johnson is embracing the concept of a bilingual employee — but not in the classic sense.

For the pharmaceutical company, literacy is needed in specialized and core job skills, including research, supply chain, and finance. Then there's fluency in AI technology.

"There are so many ways we've been using AI," said Jim Swanson, the chief information officer of J&J. "But to do that effectively, we had to really create a curriculum and a mindset around upskilling."

More than 56,000 of J&J's 138,000 workers have taken a generative AI training course, which is required before any employee is authorized to use the technology. After training, J&J's employees can utilize generative AI tools for summarization and prompt engineering, the latter a skill to ask the right question to get the best output from a large language model. A separate, more in-depth digital boot camp that covers topics including AI, augmented reality, and automation has recorded more than 37,000 cumulative hours of training from more than 14,000 employees.

Generative AI offers the promise of more quickly identifying compounds for new treatments and vaccines, accelerating drug development, streamlining regulatory compliance, optimizing which patients are best suited for clinical trials, and improving how new drugs are marketed.

Deborah Golden, Deloitte's US chief innovation officer, said these advancements were poised to change which skills the pharmaceutical industry prioritizes in recruitment. Biology and chemistry knowledge will still be needed, but it isn't as essential for newer roles like AI engineers, and other new roles might require a mix of traditional expertise and AI know-how if AI-driven drug discovery proliferates.

"When you think about how AI is shifting the balance and the talent requirements, you really need to be able to speak both the language of biology and AI models," Golden said.

How AI is changing drug development

Generative AI could save the pharmaceutical industry tens of billions of dollars each year through improved productivity within drug development.

J&J, the maker of treatments like the immunosuppressive drug Stelara and Darzalex, a medication for treating the cancer multiple myeloma, has used more traditional forms of AI for almost a decade. Use cases include AI-enabled software tools that can guide a surgeon through a procedure, speed up drug discovery, and help drug makers manage inventory more effectively.

In 2023, J&J piloted a six-week digital immersion program that focused on AI, data science, and other emerging technologies. More than 2,500 employees participated last year, taking 90-minute classes each week, and J&J is planning further expansion this year.

Swanson told Business Insider it was critical for company leaders to create a culture that promotes technological literacy. "We've been around 135 years. We've had to reinvent ourselves multiple times to stay relevant and current," he said.

The pharmaceutical giant Merck's early generative AI investments included the development of a proprietary platform called GPTeal. Merck — which is responsible for the HPV vaccine Gardasil and the immunotherapy drug Keytruda — said that GPTeal gives employees access to large language models such as OpenAI's ChatGPT, Meta's Llama, and Anthropic's Claude while keeping company data secure from external exposures.

Employees are also using generative AI to draft emails and memos and for other productivity-focused tasks, but Merck's aspirations are also getting bolder.

"Now, the journey is clearly to identify, implement, track, and measure use cases that have a dramatic impact on our business," said Ron Kim, a senior vice president and the chief technology officer of Merck.

Generative AI allows Merck's employees more time to focus on higher-impact tasks. In drug discovery, for example, generative AI can help draft (human-reviewed) regulatory documents that are submitted to health authorities. "We felt like some of our scientists were taking time being copyeditors," Kim said. "That's not what they trained for."

Kim said more than 50,000 Merck employees were using GPTeal regularly. The company supported upskilling through a mix of self-serve digital training courses, monthly webcasts focused on generative AI, and boot camps for software developers that could last anywhere from half a day to 10 days.

AI can appeal to pharmaceutical companies of various sizes

Dr. Daniel Stevens, the chief medical officer at Blue Earth Therapeutics, said AI was alluring to the clinical-stage radiopharmaceutical company because, as a small startup founded in 2021, it has to be judicious with how it spends capital.

"The application of artificial intelligence is of interest, because it may help us with some of our efficiency goals," Stevens said.

A $76.5 million Series A in October, which included funding from the healthcare investment firm Soleus Capital and the diagnostic imaging company Bracco Diagnostics, was mostly intended to support clinical trials that will assess the safety and effectiveness of new prostate cancer treatments.

Stevens said that with just 20 full-time employees, Blue Earth has not yet needed to offer AI upskill training. He added that when Blue Earth grows its employee base and is ready to offer instruction about the technology, it plans to use online courses and AI certifications from external vendors.

Eli Lilly, the pharmaceutical giant behind treatments including the antidepressant Prozac and the type 2 diabetes and weight loss medicine Mounjaro, has used generative AI to support the research of both small and large molecules. The company also used AI to generate documentation for clinical trials and create materials for regulatory submissions.

After ChatGPT launched, major employers such as Apple and Amazon restricted employee use of the popular chatbot, with many citing concerns about data privacy. "We went in the exact opposite direction," said Diogo Rau, the chief information and digital officer at Eli Lilly.

Rau encouraged Eli Lilly's workforce to embrace the tool without exposing sensitive company information, similar to how an employee might use Google Search.

"We told everybody you need to use it, you need to start bringing ChatGPT into your work," said Rau. But, he added, "Don't put anything in there that you don't want to get out."

The company also internally sought to bolster interest with an "AI Games" competition timed to the Summer Olympics in Paris. Contests involved using a chatbot to write a message to a colleague or relying on generative AI to make a quiz about Eli Lilly's history.

In 2024, Eli Lilly also encouraged all employees and managers to use generative AI for their year-end reviews. This year, the company is set to require all senior leaders and managers to obtain an AI certification.

"We've got a workforce that is embracing AI," Rau said, adding that employees often stopped him in the office or emailed him to share the ways they were applying AI to their daily work tasks.

• In a Deloitte survey, 26% of leaders said their organizations were seriously exploring autonomous agents.
• ServiceNow, SAP, and Salesforce are among the firms that have debuted AI agents to do work tasks.
• This article is part of "AI in Action," a series exploring how companies are implementing AI innovations.

When clients of the cloud-based-software provider ServiceNow contact the company's customer support center, 80% of the cases — in the form of calls and chat messages — are handled without any human intervention.

Instead, the company relies on analytical and generative artificial intelligence — in the form of AI agents — to address common customer questions.

Chris Bedi, ServiceNow's chief customer officer and enterprise-AI advisor, said employees still handle one out of every five customer-support requests.

They're getting new support from agentic AI, which can automate tasks such as drafting a response email to a customer. Workers remain in the loop for a final sign-off before any agentic-AI actions are executed. The combination of human workers and agentic AI shrank the amount of time it took to handle the more complex cases by 52% in a two-week period, ServiceNow said.

OpenAI's cofounder Sam Altman and other leading technologists have said that 2025 will be the year that AI agents "join" the workforce.

In addition to ServiceNow, software developers such as Salesforce and SAP have rolled out their own agentic-AI platforms. These can perform workplace tasks such as processing customer invoices, providing customer support to clients, and drafting emails. The business software giant Intuit, which owns TurboTax and QuickBooks, began rolling out agentic-AI capabilities in December.

Humans mostly remain in the loop for now, but vendors anticipate this technology will become fully autonomous. Multiagent systems, where two or more AI agents collaborate to complete work, will proliferate.

"Agents are the next level of understanding around how you apply AI," Jim Rowan, the head of AI at the consultancy Deloitte, said. "It can perform actions for you."

In a recent Deloitte survey of 2,773 business leaders, 26% of respondents said their organizations were exploring autonomous agents to a "large or very large extent."

Why AI agents have become the new focus for generative AI

For the first two years of the generative-AI boom — which kicked off after the debut of OpenAI's ChatGPT in late 2022 — most businesses that adopted the technology scaled it to power chatbots and complete routine tasks like drafting meeting summaries. AI agents represent an evolution of generative-AI technology, built to complete tasks autonomously, though most are still monitored closely by workers.

Agentic AI "actually possesses some unique skills around reasoning, planning, and orchestration," Bedi told Business Insider. "These agents can collaborate with each other and really start to deliver on the promise of work happening autonomously."

Buzz for AI agents kicked into high gear after Salesforce debuted Agentforce in September to automate tasks in customer support, sales, and marketing. The company has said it will roll out 1 billion agents to customers by the end of this year. The company also reported that more than 340,000 of its customer support questions had been answered autonomously with Agentforce.

ServiceNow estimates that the company's AI agents, already deployed in various parts of the business, such as customer service, human resources, and IT, are driving an estimated $325 million in annualized value by bolstering workplace productivity by 20%. ServiceNow says AI-agent-supported work saves 400,000 labor hours annually.

Still, technology companies are in the early stages of their agentic-AI development. Many are figuring out which processes they can fully automate with the technology. As a result, company leaders implementing agentic AI are training their workers to collaborate with — and provide feedback on — their new "coworkers."

AI agents are often developed as worker-collaboration tools

John Kucera, the senior vice president of product management at Salesforce, recommended that businesses be transparent about what work AI agents can handle and what will remain with workers. He added that businesses should be clear about what an AI agent actually is, saying that not all agentic systems are equal.

"There's a lot of false agents out there," Kucera said. "It's only an agent when you're taking a request and the agent is figuring out what to do and then what data to put in."

While surveys frequently find that many workers worry that AI will replace them, technologists say AI agents won't replace people but assume responsibility for mundane tasks.

"These agents are going to help me do my job, but at no point will they make me do something I'm not aware of," said Walter Sun, the global head of AI at SAP, which sells software for financial, supply chain, and other business management needs. "The most important thing is that the employees are always in control."

How companies are tailoring AI agents with employee feedback

To ensure workers have a voice in how AI agents are developed, SAP encourages employees across its various business lines — including the travel- and expense-management provider Concur and SuccessFactors, which provides HR, payroll, and talent management software — to use an internal online form to reach out to the AI team and propose compelling agentic use cases.

At Intuit, the AI-powered financial assistant Intuit Assist can get businesses paid 45% faster by detecting past-due invoices and automatically drafting a personalized reminder note. After a business owner approves the note's language and sends it out, they are paid, on average, five days sooner than with a human-only process, Intuit Assist said.

But before Inuit Assist takes action, humans have the final say. "What we're trying to do is have the right human-automation interaction," Ashok Srivastava, Intuit's chief data officer, said.

Intuit has embraced a robust AI-training program, focused on responsible AI and what the technology can and cannot do, and built a "sandbox" called GenStudio that allows employees to interact with large language models in a secure environment. The company has also developed educational programs tailored to senior executives, directors, and engineers. "It's very pervasive across the company," Srivastava added.

Asana, which makes work-management software, launched AI agents in October, focusing on a few functions, including marketing, IT, HR, and research and development. Rather than track a specific number of actions that agentic AI takes over, Asana monitors the types of work that can be automated, eliminating the drudgery of busy work to allow employees to focus on more complex tasks.

The company is also keeping a close eye on which tasks AI systems get wrong compared with people. In cybersecurity, human errors tend to occur later in the day, when workers are tired after a long shift. AI doesn't get tired, but it is susceptible to hallucinations — or when an AI model generates a response that is misleading or false information but nonetheless presents it as fact. For example, Asana's AI agent might respond to certain questions by suggesting tasks that are, in reality, nonexistent to a particular workflow.

"The kinds of errors we see are different, so the way we fix them needs to be different," Asana's chief information security officer, Sean Cassidy, said. He said the company conducts automated tests to detect hallucinations and improve the product when they occur.

For AI agents to create a compelling return on investment for the companies that implement them, Deloitte's Rowan said, workers should be tasked with frequently checking on simple automated tasks before any agentic AI actions are taken.

If businesses want to see big returns on their agentic AI investment, they need to place AI at the center of their work model, and then consider how humans will engage with the work, Rowan said. If not, "the savings really won't be there," he added.

ServiceNow's Bedi said the success of AI agents depends on companies nailing three factors: New agentic capabilities should be developed for each department and its specific needs, unique training plans should be designed for every part of the business — like finance, marketing, and sales — and the value and return on of agentic workflows should be closely monitored.

"The companies that combine all three of those ingredients are going to have a competitive advantage," Bedi said.

• Healthcare-focused AI startups are raising billions to help improve the US system.
• AI can help streamline clinical documentation, drug research, and medical billing.
• This article is part of "Trends in Healthcare," a series about the innovations and industry leaders shaping patient care.

The founder of Suki, a startup that uses artificial intelligence to automate healthcare documents, raised $70 million in funding from investors in a Series D round that was disclosed this past fall.

He said it really didn't take that much persuading: With an epidemic of stressed- and burned-out physicians, there was an obvious need for their AI software, he added.

"Most of the investor conversations over the last year and a half have been, 'Well, it looks like the market is here,'" said Punit Singh Soni, Suki's founder. "Are you going to be the winner or not?"

Suki sells an AI-powered assistant that takes notes during a conversation between patients and clinicians. The notes can be reviewed by the doctor and submitted as an electronic health record. This saves time on administrative tasks and allows physicians more time to take care of patients, a resource that's becoming increasingly limited among healthcare professionals.

Surveys have consistently found that doctors and other medical workers are burned out from working in an often overloaded, convoluted, and inefficient system. The US spent $4.8 trillion on healthcare in 2023, according to a January report from the Peter G. Peterson Foundation. The US also spends more per person than nearly all other developed nations, according to a report by the Organization for Economic Co-operation and Development. Despite this, health outcomes were poorer, with Americans facing a lower life expectancy, higher rates of treatable and preventable excess deaths, and less efficient healthcare systems.

Cash-strapped hospitals and private practices have lagged behind the financial-services and telecommunications industries in applying newer technologies, but the healthcare industry is increasingly considering artificial intelligence as it contends with high labor costs and a lot of opportunities to automate routine tasks. The pandemic exacerbated these challenges with staffing shortages as overworked doctors and nurses quit the profession.

To make healthcare more efficient, AI startups like Suki, Zephyr AI, and Tennr have raised millions with vast promises, including making repetitive tasks like billing and note-taking easier, improving the accuracy of clinical diagnosis, and identifying the right patient population for emerging treatments.

But the challenges are vast. The healthcare industry's budget allocations for generative AI are trailing those of many other core industries, such as energy and materials, consumer goods, and retail. Clinical diagnosis will continue to require a human in the loop, so the process can't be fully automated. The healthcare industry is highly regulated, and quite often, venture capitalists will wait for clarity on laws from the federal government before aggressively pushing AI tech advancements forward.

A $370 billion bet on boosting the healthcare sector's productivity

The consulting firm McKinsey estimates that generative AI can boost productivity for the healthcare, pharmaceuticals, and medical-products industries by as much as $370 billion by accelerating drug research, making clinical documentation easier, speeding up medical billing, and helping doctors make diagnoses.

Some big funding rounds announced in 2024 highlight the diverse use cases for AI in the healthcare sector. They include $150 million raised by the clinical-documentation AI startup Abridge in February, the drug-discovery AI startup Xaira Therapeutics bringing in $1 billion before its launch in April, Atropos Health's $33 million Series B in May to help doctors analyze real-world evidence with generative AI, and the medical-billing-automation provider Candid Health raising $29 million in September.

Parth Desai, a partner at Flare Capital Partners, has steered investments into healthcare startups such as Photon Health and SmarterDx. He said that healthcare organizations had been dedicating more money to bolster their AI strategies, beginning in late 2022 and accelerating through 2024. That's boosting demand for the tools these startups are developing. There's also less pressure to immediately prove a return on investment, which budget-conscious health systems have closely monitored in the past when allocating dollars for technology.

"The thing that we're really studying before making an investment decision is: Do budgets exist today to pay for this technology?" Desai told Business Insider. "Or are they going to exist in a large-enough fashion in the next five to 10 years to support this technology?"

Candid Health and Akasa aim to cut costs and automate medical billing

One area of particular promise has been medical billing, which could benefit from large language model automation. An LLM could, for example, analyze a large volume of claims in a client's system and accurately match them with insurers' unique billing codes, a process required for repayment to a physician for their services. Hospitals have traditionally relied on human medical coders to hunt down reimbursements from insurers.

"The software used to do billing was built a long time ago and basically wasn't kept up to date," Nick Perry, a cofounder and the CEO of Candid Health, said.

Malinka Walaliyadde, the CEO of Akasa — another medical-billing-focused AI startup — said the company builds customized LLMs for each healthcare institution it serves. Typically, the aim for these LLMs is to lower costs by lessening the reliance on human medical coders. This often reduces errors in billing and speeds up repayment cycles.

"We looked at what are the biggest pain points for health systems," Walaliyadde told BI. He said that Akasa's focus is on developing LLM products for medical coding and simplifying prior authorization, a process that requires approval from a health-plan provider before a patient can receive a treatment. "Those are the ones where you could really move the needle," Walaliyadde said.

AI for health screenings

George Tomeski, the founder of Helfie AI, is in the middle of pitching investors to raise as much as $200 million in a new round of funding that he hopes to close by the first half of 2025.

Tomeski said the funding would help Helfie scale as it exits beta testing for the company's app. The app, also called Helfie, uses a smartphone camera to do medical "checks" that screen for illnesses including COVID-19, tuberculosis, and certain skin conditions.

"We're targeting all the health conditions that lead to avoidable mortality," Tomeski said, adding that the app focuses on respiratory and cardiovascular conditions. The intention is for these checks —which can cost as low as $0.20 a person per screen — to serve as a form of preventive care and as an incentive to go see a doctor in person.

While some funding is going toward sales and marketing, talent acquisition, and ensuring adherence to regulations around privacy and healthcare data, a large chunk is still being allocated to product development as AI tech advances quickly.

Dr. Brigham Hyde, a cofounder and the CEO of Atropos Health, said his latest funding announcement, in May, was timed to coincide with the geared-up launch of ChatRWD, an AI copilot that can answer doctors' questions and quickly churn out published studies based on healthcare data. Hyde said he's keen to bring in big partners this time, including the pharmaceutical giant Merck and the medical-supplies and equipment maker McKesson.

But Hyde also had to show some restraint. He said that when Atropos Health moved forward with its Series B rounds, dozens of venture capitalists expressed interest in leading the round. The company was offered up to $100 million but took only one-third of that amount.

"I don't always think that's a good idea," Hyde told BI. "As a founder, you want to raise the right amount of money for your business and for the stage you're at."

It may be tempting to take more, as many healthcare AI startups — a vast majority still in the seed and early-stage funding rounds — are racing to outmaneuver rivals. Even if the technology is right, it has to get past regulatory approvals and persuade cautious hospitals and health systems to open up their wallets.

"You can build as much product as you want, but you can never build a market," Soni of Suki said. "It shows up, or it doesn't show up."

• Many workers struggle with choosing their health-insurance plans during open enrollment.
• Some healthcare companies are employing mobile apps and generative AI to help smooth out the process.
• This article is part of "Trends in Healthcare," a series about the innovations and industry leaders shaping patient care.

During open-enrollment season, Reddit users inundate the platform's forums on health insurance and personal finance every day, asking how to best pick from their health-insurance options.

In one post, a recently unemployed married woman in Texas asked whether she should enroll with her husband's employer or stick to COBRA, which provides benefits to people who have lost their jobs. Another married person requested advice on which coverage to pick if they're planning to have a baby in 2025. For an employee in California, fellow Redditors were a sounding board as they navigated dental-plan options, with costs ranging from $0 to nearly $440 annually.

Open-enrollment season typically takes place between October and December, and companies have their own set periods within those months. During this time, Americans elect their health-insurance coverage through either a private employer or marketplaces via subsidies offered under the federal government's Affordable Care Act. Nearly all open-enrollment selections made this fall will go into effect on January 1 and be set until the following season, with a few exceptions.

The process can be immensely confusing.

Employees are expected to look both backward and forward, said Dan Beck, the president and chief product officer for SAP SuccessFactors, a cloud-based software platform that oversees HR, payroll, and talent management.

He told Business Insider that employees are tasked with reflecting on whether they maximized their benefits in the past year based on how much they tapped into the healthcare system. At the same time, they must anticipate health-related events, such as having a child or a major surgery.

To complicate matters further, workers may move to new roles with different insurance options or their employers could change providers or plan options, forcing employees to acquaint themselves with new choices. The makeup of their families could change, too: As employees' marital statuses change and they raise children, they'll likely want to optimize their healthcare plans for those life stages.

On Reddit, people making health-insurance decisions try to make sense of the complexities. If they choose health plans mismatched with their needs, they run the risk of overspending in two directions: shelling out for a premium-coverage plan they don't really need or skimping on coverage and then experiencing an expensive life change.

Employees also need to keep track of life events that could change their coverage, including moving, having a baby, or adopting a child. There is a special enrollment period, outside of open enrollment, for those life events, but also a limited period of time to make the changes and retain health care coverage.

Increasingly, employers are encouraged — by both their employees and their HR-benefits companies — to share more easily digestible benefits information.

"What employees are telling us, overwhelmingly, is that they need help when they are enrolling," Karen Frost, a senior vice president at the cloud-based employee-benefits vendor Alight, told BI.

Some employers are partnering with third-party companies that handle things like payroll and health benefits and have built software with with clear step-by-step prompts — which can help workers be confident about their healthcare elections.

Young workers want more employer support in demystifying healthcare

In Alight's 2024 annual survey of 2,500 employees in the US, the UK, France, Germany, and the Netherlands, 63% of workers said they felt confident about their most recent health-plan election.

There are, however, some generational splits in the data. In the survey, 70% of Gen Z and 72% of millennial workers said they wanted personalized support for navigating the health system versus just 46% of baby boomers.

Before the mass digitization of benefits elections, employers would hand their employees printed packets outlining their medical-insurance offerings and ancillary benefits such as dental and vision, retirement plans, commuter reimbursement, gym memberships, and other wellness programs. Employees would pick from that menu, largely without guidance or input on what they'd like to see as alternative or additional benefit options.

Though the paper-packet method is much less common now, the enrollment process can still be overwhelming to navigate.

Life changes, like moving to a new state or employer, or a company picking new insurance providers to work with can complicate the enrollment process.

"You have a narrow window to actually get benefits, and you want to be successful," Beck told BI.

Mobile apps and generative-AI tools aim to smooth out the open-enrollment process

To help employees sort through their options during the open-enrollment period, some healthcare startups are leveraging mobile apps and generative-AI chatbots.

Alight, for example, aims to learn more about employee preferences and the needs of their families through a Q&A and then make recommendations. Throughout this process, Alight's recommendations coincide with clear definitions of complex benefits, like a health savings account, which lets workers set aside pretax money for qualified medical expenses.

"Instead of just letting people make their own choice, we guide them," said Frost. As an example, if an employee were to pick a high-deductible health plan, Alight would guide them to an HSA and explain why enrolling in it may make sense to budget for potential healthcare expenses.

SAP SuccessFactors said it's not yet comfortable with offering suggestions for health-insurance elections, citing concerns about data privacy.

Instead, the company — which has customers including McDonald's, L'Oréal, and Delta Air Lines — said it's focusing on further developing its recently launched mobile app.

The SuccessFactors mobile app is targeted at two demographics: workers under 40 who tend to be mobile-first in nearly every aspect of their lives and frontline workers of all ages with jobs in manufacturing and other sectors where they may be without frequent access to computers.

SAP SuccessFactors is also using generative-AI chatbots to answer policy questions to improve the user experience. In the future, the company plans to use these chatbots to automate some open-enrollment processes.

To bolster the company's abilities to help employees navigate this process and other healthcare questions that may arise throughout the year, SAP earlier this year paid $1.5 billion in cash to buy WalkMe, a tool designed to provide real-time website navigation for healthcare, onboarding, and other employee-focused tasks.

AI-based virtual assistants are also becoming more pervasive in the open-enrollment process. Alight has Ask Lisa, SAP SuccessFactors is leaning on the company's artificial-intelligence copilot, Joule, and the HR- and financial-software provider Workday uses Wex, an AI chatbot that internal employees can access on Slack to get automatically generated responses to their benefits questions. The same tool is offered to customers but branded as Workday Assistant.

"We try to appeal to all generations and age groups," Ben Carter, the senior vice president of business partners and rewards at Workday, said. "Some people, the last thing they want to do is actually talk to somebody on the phone."

This emerging technology benefits employers, too. Earlier this year, Workday unveiled an AI-enabled tool called Workday Wellness, which integrates with insurance providers like Aetna and Cigna. It allows Workday's customers — like The Hartford, Guardian, and MetLife — to understand which wellness benefits employees are using and which ones aren't resonating so they can invest more strategically.

"It brings a nice story," Carter said, "to say, well, if I'm going to go invest another $20 million in my benefits programs next year, here's where I need to go, or here's where I need to double down, or here's where I need to stop investing."