OpenAI has raised its maximum security bounty to $100,000 and introduced a five-tier loyalty program for researchers. The company also launched a security sprint with Cathedral to stress-test core products like ChatGPT and Codex. Vulnerabilities in these systems now pay up to $30,000, while general findings can earn researchers up to $172,000 with bonuses.
The firm retroactively paid the difference in bounty rates to researchers who reported vulnerabilities in the last 30 days. Exceptional findings have a potential cap of $100,000, and the new loyalty program offers additional rewards for quality and consistency. OpenAI is actively seeking external help to identify critical flaws in its systems.
This aggressive bounty increase signals OpenAI’s urgency to secure its products against emerging threats. The security sprint with Cathedral will likely uncover hidden weaknesses before they are exploited. Developers and companies relying on OpenAI’s tools should monitor these updates closely, as a breach could have significant implications for their own operations.
