• guardrails — The checks around a model that block bad inputs and outputs — filters, validators, and permission rules the model itself can't override.
• AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.

Anthropic holds AI-written production code to a higher bar than human code, using layered automated checks including Claude-powered fuzzing and automated security review. It's a concrete blueprint for what guardrails an org running agentic coding at scale actually needs.

• articleQuoting Jakub Pachockisimonwillison.net
• articleResearch acceleration: The view inside OpenAIsimonwillison.net
• articleFeeling sad about AIsimonwillison.net

Checking sign-in…

Loading comments…