Google has introduced a new bug bounty program specifically targeting security vulnerabilities in its AI products, offering rewards of up to $30,000 for critical discoveries.
What Qualifies for Rewards?
The program focuses on serious security flaws that could lead to:
- Unintended command execution (e.g., unlocking doors through Google Home)
- Data breaches (such as sending user emails to attacker-controlled servers)
- Other significant security compromises in AI systems
Important: Simply inducing the AI to generate incorrect responses does not qualify for compensation.
Reward Structure
- Standard maximum: $20,000
- Enhanced payout: Up to $30,000 for original, high-quality reports
This initiative reflects Google’s commitment to proactive AI security, encouraging researchers to identify and report vulnerabilities before they can be exploited.
The program targets real-world threats that could compromise user privacy and home security systems integrated with Google’s AI products.
Source: The Verge
