Claude Updates

Former Anthropic researcher Jacob Coxon expressed concerns that artificial intelligence systems could eventually become sophisticated enough...

Researchers conducted a study testing whether Claude could be used to develop biological weapons, examining the AI system's safety measures...

Anthropic has alleged that Chinese AI companies Moonshot and DeepSeek redirected user requests to Claude without disclosure, raising concern...

Anthropic has published research documenting attempts by malicious actors to use its AI models to develop biological weapons, revealing spec...

Two AI researchers have departed from Anthropic and Google citing concerns that leadership is not adequately prioritizing AI safety measures...

Anthropic researchers have outlined three distinct scenarios for how artificial intelligence could reshape the economy and knowledge work, e...

Anthropic declined requests to use Claude for developing biological weapons, according to the company's safety practices.

Anthropic reported discovering and blocking an attempt to use its Claude AI model to assist with developing biological weapons, which the co...

Anthropic reported identifying and blocking efforts by users attempting to use its AI systems to advance biological weapons development, acc...

Anthropic disclosed that researchers used its Claude AI system to conduct experiments potentially applicable to biological weapons developme...

Anthropic published research on methods for identifying and mitigating harmful applications of AI systems as of September 2026.

US lawmakers are pushing for new artificial intelligence regulations following recent warnings from Anthropic researchers about AI safety ri...

Anthropic reported detecting and blocking attempts to use its AI systems to research biological weapons development.

Researchers from OpenAI and Anthropic have called for a deceleration in AI development following recent departures from both companies.

Anthropic has released an interactive tool that models various economic scenarios involving artificial intelligence, allowing users to explo...

Elon Musk criticized warnings about AI extinction risks from a former Anthropic employee, suggesting the concerns were contrived. The exchan...

Anthropic provided the European Union's cybersecurity agency with access to its Mythos 5 AI model for security testing and evaluation purpos...

The European Union gained access to Anthropic's most advanced AI model approximately three months after its initial release, following regul...

Several researchers from major AI labs including OpenAI and Anthropic have added their voices to calls for slowing AI development, citing co...