OpenAI wants to ‘have their cake and eat it too’ with health AI policy recommendations

May 6, 2026 - 12:18

OpenAI has released a set of policy recommendations for using artificial intelligence in health care, and experts say the proposals are a mixed bag. The blueprint, titled "A Blueprint for Unlocking AI's Potential to Change Health Care," outlines how the company thinks regulators should handle the fast-growing use of AI in medicine. While some of the ideas are sensible, critics argue the document is designed to benefit OpenAI more than patients or the public.

The recommendations include calls for clearer FDA guidelines, more federal funding for AI research, and liability protections for developers when their tools are used by doctors. These are reasonable requests on the surface, but they also shield companies like OpenAI from legal risk while pushing for fewer restrictions on their products. One health policy expert described the approach as wanting to "have their cake and eat it too," noting that the company seeks the credibility of being regulated without the full burden of accountability.

Another sticking point is the suggestion that AI systems should be treated as medical devices only in certain cases. This could allow OpenAI to sidestep the most rigorous testing requirements for some of its tools. Critics also point out that the blueprint focuses heavily on what regulators can do for AI companies, with less attention to patient safety, data privacy, or the risk of biased algorithms.

The timing of the document is notable. As federal agencies scramble to create rules for AI in health care, companies like OpenAI are trying to shape those rules early. The result, experts say, is a policy wish list that sounds helpful but ultimately serves the bottom line.