Google Launches AI Bug Bounty Program, DeepMind Unveils CodeMender
Google has launched an AI bug bounty program to identify vulnerabilities in its AI products, including Bard, Gemini, and Assistant. The program is an expansion of its existing Abuse Vulnerability Reward Program (VRP) and offers rewards of up to $30,000 for particularly dangerous exploits.
Since its inception, the AI bug bounty program has seen external researchers earn over $430,000 in rewards. Google retains 'bonus multipliers' from the previous program, allowing a single report's reward to reach up to $30,000. The program offers up to $20,000 for well-written error reports in flagship AI products like search, Gemini, Gmail, and Drive.
Meanwhile, DeepMind has announced an AI agent called CodeMender. This agent proactively rewrites existing code in open-source projects to use safer data structures and APIs. In the past six months, CodeMender has made 72 security fixes in open-source projects, including those with up to 4.5 million lines of code. The AI bug bounty program also considers qualified errors such as invisible prompt injections that change account or associated product status.
Google's AI bug bounty program encourages external researchers to help improve the security of its AI products. DeepMind's CodeMender demonstrates the potential of AI in proactively enhancing the security of open-source projects. Both initiatives highlight the growing role of AI in cybersecurity.
Read also:
- Trump and Xi speak over the phone, according to China's confirmation.
- NVIDIA introduces Blackwell to the cloud and unveils the significant enhancement of GeForce Now at Gamescom 2025, marking a major step in cloud gaming technology.
- Strategies for Adhering to KYC/AML Regulations in India, a Leading Fintech Center (2024)
- Strategies for Poland, Ukraine, and NATO to combat unmanned Russian aerial threats.
