Agents
154d agoOpenAI Launches Safety Bug Bounty
OpenAI has announced a global Safety Bug Bounty program aimed at uncovering agentic vulnerabilities, prompt injection, and data exfiltration risks in its AI systems. By inviting researchers worldwide to test and report flaws, the initiative seeks to strengthen AI safety and build public trust.
The Signal
The program underscores the importance of proactive risk mitigation in advanced autonomous systems.