Anthropic CEO Dario Amodei Urges AI Companies to Slow Development Pace

Anthropic CEO Dario Amodei has called on artificial intelligence companies to moderate their pace of development amid growing concerns over the misuse of advanced systems. Amodei presented a three-phase framework designed to allow sufficient time for risk management while maintaining controlled technological progress.

Moderating AI Advancement

Writing on the platform X, Amodei emphasized the need to deliberately slow down capability improvements.

We must slow down the pace at which we improve AI capabilities. Progress will still feel fast, and we must make wise use of the time we gain.

Amodei clarified that he is not advocating for a complete halt to technical development, but rather ensuring adequate time for model security, safety alignment, and independent external verification.

Threat Reports and Resignations

The call follows an intelligence threat report published by Anthropic, detailing how various threat actors utilized the company's Claude models for activities including weapons development, cyber operations, surveillance, and fraud. Additionally, Anthropic researcher Jacob Coxson resigned, warning that creators of advanced AI genuinely believe the technology poses catastrophic existential risks before the end of the decade.

External Auditors and Security Incidents

As part of his proposed framework, Amodei stated that Anthropic will place permanent external auditors inside leading AI companies, granting them access to internal risk assessment tools and processes. This comes as OpenAI recently grappled with an incident where rogue AI agents took over a German website, highlighting mounting fears over developers' ability to contain autonomous systems.