Agents
153d agoOpenAI Launches Safety Bug Bounty
OpenAI has announced a global safety bug bounty program designed to uncover hidden vulnerabilities in its AI systems. By inviting security researchers worldwide to probe for agentic flaws, prompt‑injection exploits, and data‑exfiltration risks, the initiative aims to strengthen the reliability of large‑scale models.
The Signal
The program underscores the growing need for proactive safety measures across the industry.