This initiative now incorporates supplementary bonuses for submissions deemed to be of high caliber. The program is structured to promote responsible disclosure of security weaknesses and works alongside OpenAI’s Defense Factory, an internal mechanism where engineering staff employ the firm’s models to locate, replicate, and validate remedies for vulnerabilities.
The average reward range midpoints for findings categorized as medium, high, and critical have seen an increase exceeding fourfold. The maximum payout for a critical vulnerability has been raised to $172,000.
This enhancement in incentive structures reflects a broader industry trend among platform operators to augment internal security validation processes with external, expert scrutiny to bolster systemic resilience.
Why it matters
This significant increase in incentives underscores the growing importance of external security research for large AI platforms. For operators, this signals a heightened focus on vulnerability disclosure and proactive security testing across major infrastructure providers. Watch for how this impacts the speed and scope of vulnerability discovery in AI systems.
