Geoffrey Hinton says Anthropic has drifted from safety mission

Geoffrey Hinton, widely known as the "godfather of AI," believes Anthropic has strayed from its founding commitment to AI safety as the company faces mounting competitive and financial pressures, he told NBC News in a late May interview.

The comments represent a significant critique from one of the field's most respected voices. Hinton suggested that Anthropic, which positioned itself as a safety-focused alternative when it launched, is now pursuing goals that conflict with its original mission.

Why it matters

Anthropic has marketed itself as the responsible AI company, attracting talent and investment based on that reputation. If one of AI's most prominent researchers believes the company is compromising on safety, it raises questions about whether any commercial AI lab can maintain safety commitments when facing market pressures. The assessment also signals growing concern among AI pioneers about the industry's direction as competition intensifies.

Shift on military AI

In the same interview, Hinton revealed a reversal in his thinking about military applications of AI. He now sees a role for artificial intelligence in military operations, a position that contrasts with his previous skepticism about such uses.

The shift reflects the broader evolution in Hinton's public stance on AI risks. After leaving Google in 2023, he has become increasingly vocal about the potential dangers of the technology he helped create, while also grappling with practical questions about how AI will be deployed.

Anthropic's trajectory

Anthropic was founded in 2021 by former OpenAI executives who left over disagreements about the company's direction and safety approach. The startup has raised billions in funding, including major investments from Google and Amazon, while developing its Claude AI assistant.

The company has consistently emphasized its commitment to AI safety research and responsible development practices. However, Hinton's assessment suggests that maintaining those priorities has become more difficult as Anthropic scales and competes with well-funded rivals like OpenAI and Google.

The competitive dynamics in AI have intensified dramatically, with companies racing to release more capable models and secure market position. That environment creates pressure to move quickly, potentially at the expense of more cautious safety-oriented approaches.

These details were first reported by NBC News.