AI Reshapes Healthcare: Pfizer, Anthropic and Longevity Labs

Pfizer, Anthropic and leading aging researchers outlined recent developments this month that industry participants say are changing how drugs and medical research are done. The companies and scientists described progress in AI-driven molecule design, model development and longevity research, and highlighted technical and regulatory limits that remain.

In early May Pfizer’s chief executive said the company is reviewing a candidate molecule generated by internal AI design teams. Pfizer has paid outside groups for AI-designed compounds since 2020, invested in AI biology ventures and formed a collaboration to refine biomolecular models on its own data. The company’s executives have said AI is speeding work across discovery, with the aim of shortening the time from target identification to a candidate medicine. Pfizer Ventures has invested in longevity-focused projects as part of that wider strategy.

At an invite-only financial services event in New York, Anthropic’s chief executive described a measured lead for the company’s frontier models over other U.S. labs and estimated Chinese labs lag by roughly six to 12 months. The company released an updated model and introduced agent-style tools for banks, including a financial-crime application developed with a payments partner. Anthropic reported that one of its models flagged tens of thousands of previously unknown software vulnerabilities and warned governments and large organizations they have roughly a six- to 12-month window to patch systems before competing models narrow that gap. The company’s pre-IPO valuation topped $1 trillion in April and first-quarter revenue grew sharply on an annualized basis.

On a recent podcast, immunology professor Derya Unutmaz predicted that most diseases could be addressed within 10 to 15 years using AI-enabled drug discovery, drug repurposing and personalized therapies. Biomedical gerontologist Aubrey de Grey estimated about a 50 percent chance of reaching ‘‘longevity escape velocity’’ by the late 2030s. Unutmaz added, “Very soon it’s going to be malpractice not to use AI in medicine.”

Executives and researchers pointed to limits that will affect timelines, including available computing resources, gaps in biological data and the need for regulatory frameworks that allow testing and approval of AI-designed therapies. They noted three specific verification points: whether Pfizer’s AI-designed molecule advances to clinical trials, whether Anthropic’s lead holds up under external benchmarking, and whether longevity teams produce reproducible animal results.

Financial activity reflects private investment in AI-biotech links. Biopharma companies have paid outside groups for molecule proposals, venture arms have backed community longevity projects, and frontier AI labs have reported rapid revenue growth and high private valuations as they commercialize models for enterprise customers in finance and healthcare.

Generative models are being used to suggest candidate molecules, optimize trial designs and search biological data at scale. Company officials and researchers emphasized that laboratory validation, clinical testing and regulatory review remain required steps before computational designs become approved medicines. Industry participants expect more data and independent benchmarks in the coming months to clarify how quickly these AI-driven approaches will translate into approved therapies.