OpenAI significantly boosts bug bounty rewards and introduces tiered incentives for security researchers

OpenAI has substantially increased the rewards available through its Security Bug Bounty program to encourage independent researchers to find vulnerabilities in its software. The company is raising the midpoint of reward ranges for medium, high, and critical findings by over fourfold on average. For instance, a critical vulnerability in ChatGPT or the API Platform can now yield up to $30,000, with exceptional findings potentially earning up to $172,000 with bonuses. To further incentivize quality, the program introduces a bonus multiplier for reports that provide a reliable proof of concept demonstrating exploitability and impact. Additionally, a new 5-tier loyalty bonus system is being implemented to recognize researchers who consistently partner with OpenAI. Short-term priority campaigns will also feature dedicated bonus pools for newly launched features. Specific examples illustrate the new structure: a medium finding in ChatGPT could earn $2,000 plus bonuses, while a critical one in ChatGPT could reach $30,000 plus bonuses. Researchers who submitted reports before the announcement will have their awards adjusted to reflect the new ranges, and OpenAI's security team will be present at PROMPT2PWN in Oakland on November 20–21.