Google is offering a big cash prize up to ₹26 lakh (around $30,000), to anyone who finds serious problems (bugs) in its AI tools.
What’s the Program About?
It’s called the AI Vulnerability Reward Program (AI VRP), and it invites security researchers, ethical hackers, and AI developers to uncover serious flaws in Google’s AI-powered products like:
- Gemini
- Bard
- Gmail
- Workspace
- Search
- Drive
- NotebookLM
- Jules
What Kind of Bugs Qualify?
Google wants to find cases where its AI can be fooled into doing something wrong or dangerous things, it’s not supposed to do. Let’s break it down:
- Rogue actions: Bad or risky things the AI might do if tricked.
- Tricked into doing something harmful: Like unlocking someone’s smart home or sending private emails without permission.
- Unauthorized: Actions the AI should never do, like giving access to secret files or helping with scams.
Example: If someone finds a way to make Google’s AI reveal private info or bypass security, that’s a serious bug, and Google will pay big money to fix it.
Reward Structure
| Severity Level | Reward Amount |
| Major bugs in core products | Up to ₹17.75 lakh ($20,000) |
| Bonus for originality or impact | Additional ₹8.9 lakh ($10,000) |
| Minor bugs in other tools | ₹40,000 to ₹12 lakh ($500–$15,000) |
Google’s AI Tool to Fix Bugs
Google also introduced CodeMender, an AI agent that automatically patches vulnerabilities in open-source software. It’s already helped fix over 70 issues.
How to Participate?
Submit your findings through Google’s Bug Hunters platform. The program is open globally and aims to make AI safer and more trustworthy.
