AI could help cure most diseases in 5-10 years: Anthropic CEO

Anthropic Chief Executive Dario Amodei said artificial intelligence could help cure most human diseases within five to 10 years, as the AI company steps up its work in biology and medicine.

Amodei, in a post on X, said he believed AI had the potential to radically transform healthcare and biology, while acknowledging that such a prediction may sound unrealistic to many people, including scientists.

"I think it will actually be possible to cure most human disease in ~5-10 years," Amodei said, referring to an essay he wrote on the potential benefits of AI.

Amodei, who has a background in biology, said he wrote his essay Machines of Loving Grace because he felt the AI industry was not sufficiently communicating the technology's potential benefits.

He said much of the essay focused on AI's potential in health and biology and argued against scepticism about its ability to accelerate medical advances.

Amodei also said Anthropic was rapidly expanding its efforts in biology and medicine and that the company hoped to produce "incredible results" in the coming years.

The comments came as Amodei rejected criticism that his public messaging had focused disproportionately on the technology's risks.

He said public distrust of AI was instead rooted in a wider "crisis of trust" involving companies, governments and the technology industry.

Amodei said AI companies had so far failed to deliver on some of their promises to benefit society, and argued that tangible breakthroughs rather than marketing campaigns would be needed to restore public confidence.

"The thing that will work is actually curing cancer," he said, arguing that claims about AI's ability to transform healthcare would carry more weight when accompanied by real-world results.

Amodei cited his personal experience of losing his father to hepatitis C, years before the development of direct-acting antiviral drugs that can cure about 95% of patients with the disease.

Anthropic, which develops the Claude family of AI models, has increasingly focused on the potential applications of AI beyond general-purpose software, including scientific research and drug development.