Building safeguards for Claude
…Related content Government of Alberta uses Claude to find and fix cybersecurity vulnerabilities across government systems Since 2025, the Government of Alberta has been using Claude Code with both Opus and Sonnet…
Search
Trustworthy agents in practice
…This represents a new frontier for governance. Agents are already making real productivity gains for our customers and inside Anthropic . But the autonomy that makes agents useful also introduces a range of…
More details on Fable 5’s cyber safeguards and our jailbreak framework
…safe deployment of increasingly advanced AI models. We are building it based on our own experience in preventing misuse, and assisted by feedback from our industry partners and the government. We hope…
Announcing Anthropic's Responsible Scaling Policy
…Our RSP defines a framework called AI Safety Levels (ASL) for addressing catastrophic risks, modeled loosely after the US government’s biosafety level (BSL) standards for handling of dangerous biological materials. The…
Values in the wild: Discovering and analyzing values in real-world language model interactions
…Although our method could potentially be used as an evaluation of how closely a model hews to the developer’s preferred values, it can’t be used pre-deployment. That is, the…
Claude for Financial Services
…estimates Morningstar contributes valuation data and research analytics Palantir builds AI-driven platforms that help governments and enterprises integrate, analyze, and act on large-scale data to make critical operational decisions PitchBook…
Building AI for cyber defenders
…We hope to enable models to find insecure code before deployment and to find and fix vulnerabilities in deployed code. There are, of course, many more critical security tasks we did not…
Mapping AI-enabled cyber threats: Insights from the LLM ATT&CK Navigator
…Now consider an actor with explicit malicious intent who misuses Claude to develop working malware, but we have no evidence of deployment or downstream impact—yet. The multiplicative model would, again, zero…
How we contain Claude across products
…The risk of these deployments has two components: how likely a failure is, and how much damage one could do. Progress on safeguards and model training has steadily driven down the first…
Core views on AI safety: When, why, what, and how
…In addition to our research—across safety, capabilities, and policy—these concerns drive our approaches to corporate governance, hiring, deployment, security, and partnerships. In the near future, we also plan to make…