Microsoft Tries to Catch Up in AI With Healthcare Push, Harvard Deal

Microsoft has a lofty goal: to become an artificial-intelligence chatbot powerhouse in its own right rather than leaning on its partnership with the ChatGPT maker, OpenAI.

In an effort to steal a march on its more-advanced rivals, the company has seized on healthcare as a lane in which it believes it can deliver a better offering than any of the other major players and build the brand of its Copilot assistant.

A major update of Copilot scheduled for release as soon as this month will be the first to reflect a new collaboration between Microsoft and Harvard Medical School, people familiar with the matter said. The new version of Copilot will draw on information from the Harvard Health Publishing arm to respond to queries about healthcare topics. Microsoft will pay Harvard a licensing fee, one of the people said.

In an interview, Dominic King, vice president of health at Microsoft AI, declined to discuss the arrangement with Harvard but said the company’s aim is for Copilot to serve answers that are more in line with the information users might get from a medical practitioner than what is currently available.

“Making sure that people have access to credible, trustworthy health information that is tailored to their language and their literacy and all kinds of things is essential,” he said. “Part of that is making sure that we’re sourcing that material from the right places.”

King said the intent is to help users make informed decisions about managing complex conditions such as diabetes. In the past, experts have warned about relying on chatbots for medical advice. A 2024 study led by researchers at Stanford University found that out of 382 medical questions posed to ChatGPT, the chatbot gave an “inappropriate” answer on roughly 20% of them.

While the Harvard Health Publishing literature includes mental-health material, Microsoft declined to say how the updated Copilot would handle such questions. The issue of how chatbots interact with people who experience mental illness has drawn scrutiny from lawmakers and health experts, with ChatGPT playing a role in crises that have ended in hospitalization or death, The Wall Street Journal has reported.

Another tool in development would allow Copilot to help users find healthcare providers near where they live based on their healthcare needs and insurance coverage.

Mustafa Suleyman , chief executive of Microsoft AI, has made healthcare a focus of the division’s efforts as he has increased staffing at an internal AI lab that competes with OpenAI. In June, the company, which employs clinicians, said an AI tool it developed can diagnose disease at a rate four times more accurate than a group of doctors and do so at a fraction of the cost.

Despite a tentative agreement announced last month to extend the partnership between Microsoft and OpenAI, there remains urgency within Microsoft on building up a measure of technological independence from OpenAI, the people familiar with the matter said. Last week, Microsoft Chief Executive Satya Nadella said he would hand off some duties to a deputy so he could concentrate on the company’s biggest AI bets .

Microsoft, which created its division dedicated to consumer AI and research in 2024, is training its models with the goal of eventually replacing workloads from OpenAI, the people said. Achieving that might take years. The company has said that OpenAI “will continue to be our partner on frontier models” and that its philosophy is to use the best models available.

Microsoft trails OpenAI in consumer AI. The Copilot smartphone app has been downloaded 95 million times, while ChatGPT has been downloaded more than a billion times, according to Sensor Tower, a research firm.

Copilot exists as both a consumer app and a virtual assistant within the company’s enterprise tools. It currently relies heavily on OpenAI’s models to respond to queries.

Microsoft in August said that it had started to test publicly a homegrown AI model that could be used for its Copilot chatbot. Researchers and engineers , especially those recently hired away from Google’s DeepMind AI lab, are focused almost entirely on advancing Microsoft’s own models. Already, Microsoft is using non-OpenAI models for some of its other software. It now deploys models from OpenAI’s rival Anthropic to power AI tools within its 365 products.

AI has been a revenue driver for Microsoft largely through its Azure cloud-computing unit, which OpenAI and other companies use for AI computing workloads and training. It also offers AI tools in its productivity and enterprise software products, which have millions of customers.

Under Microsoft and OpenAI’s tentative agreement in September, Microsoft would potentially receive a 30% stake in a new for-profit entity OpenAI is seeking to create. The agreement isn’t yet final.

Write to Sebastian Herrera at sebastian.herrera@wsj.com