Google has launched a new AI Vulnerability Reward Program (VRP) to incentivize researchers to report security vulnerabilities in its AI systems, building on previous extensions. The program offers rewards of up to $20,000 for critical security issues across Google’s AI products, focusing on vulnerabilities like data exfiltration and account manipulation. #GoogleAI #AIVulnerabilityRewardProgram
Keypoints
- Google’s AI VRP aims to reward researchers for discovering security vulnerabilities in its AI products.
- Prompt injections, jailbreaks, and alignment issues are outside the scope of the program but still encouraged to report as content issues.
- Attacks involving data leaks, account modifications, and denial-of-service are within the scope of the VRP.
- Rewards vary from $10,000 to $20,000 depending on the product tier and severity of the vulnerability.
- A unified rewards panel will determine the highest payout for each valid security report.
Read More: https://www.securityweek.com/google-offers-up-to-20000-in-new-ai-bug-bounty-program/