Project Pilot: Can AI models fly drones?
…Being able to do this opens up a large surface over which AI could contribute to the economy, but likewise opens up a new area of risk. A key reason why Anthropic…
Search
Clio: Privacy-preserving insights into real-world AI use
…Enhanced monitoring for high-stakes events Clio also helps us monitor novel uses and risks during periods of uncertainty or high-stakes events. For example, while we conducted a wide range of…
Investigating three real-world incidents in our cybersecurity evaluations
…to the same security standard as any other system our models run in. A cyber evaluation range that contains only fictional scenarios may appear to be low-risk—after all, nothing inside…
A “diff” tool for AI: Finding behavioral differences in new models
…the most subtle risks in new models. Auditing a new model from scratch is like being handed a million lines of code and told to “find the security flaws.” It’s an…
Focus areas for The Anthropic Institute
…Research on the societal areas most in need of investment in resilience in the face of new AI-enabled security risks. More detailed information about how our work at Anthropic has sped…
Claude Fable 5 and Claude Mythos 5
…Releasing a model this capable comes with risks. Without safeguards, Fable 5’s capabilities in areas like cybersecurity could be misused to cause serious damage. We’ve therefore launched the model with…
Measuring LLMs’ ability to develop exploits
…It was developed as a collaboration between UC Berkeley, the Max Planck Institute for Security and Privacy, UC Santa Barbara, and Arizona State University (with contributions from security researchers at Anthropic, OpenAI…
Agentic misalignment: How LLMs could be insider threats
…Risks considered but outweighed: - Violation of corporate confidentiality justified by higher national security imperative Similar observations across models reveal that their harmful behaviors in these scenarios emerge primarily not due to confusion…
Introducing Claude Sonnet 4.5
…Claude Sonnet 4.5 reduced average vulnerability intake time for our Hai security agents by 44% while improving accuracy by 25% , helping us reduce risk for businesses with confidence. Claude Sonnet 4…
Measuring LLMs' impact on N-day exploits
…If you’re interested in helping us with our efforts, we have job openings available for research scientists and engineers, threat investigators, policy managers, offensive security researchers, security engineers, among many other…