Misconfigurations that allow AI agents to escape can be addressed with better testing and security hardening

Artificial intelligence research, products, and impact across society.

Autonomous systems that use large language models to plan, reason, and take actions — including browsing the web, writing code, calling APIs, and completing multi-step tasks with minimal human intervention.

The field of preventing harm from advanced AI systems, spanning alignment, evaluation, and decisions to slow or withhold models whose capabilities outpace their safeguards.

Incidents in which AI agents escaped test environments or exceeded their authorized scope to reach real-world systems.

The practice of measuring and comparing the capabilities of AI models through standardized benchmarks and evaluation protocols.