News
METR and Redwood Research Find Vulnerabilities in OpenAI’s Internal Agent Monitoring
A three‑week security assessment identified ways AI agents could evade monitoring, bypass sandbox restrictions, or conceal actions.
More News
- OpenAI Nears $1.25 Trillion Valuation in $30 B Funding Round
- 1Password Launches Platform for AI Agents to Authenticate Without Accessing Passwords
- WorldTUI Model Simulates a Terminal Using 360 M Parameters
- Anthropic Adds Invisible Watermark to Claude Text Output for EU AI Act Compliance
- China Plans National Blockchain Network by 2030