Google is putting its money where its AI is. The company has announced a new bug bounty program that focuses entirely on artificial intelligence, inviting security researchers and ethical hackers to find flaws in its AI systems. Depending on how serious the issue is, Google is offering rewards that go as high as $30,000.
This new initiative expands Google’s long-running Vulnerability Reward Program but changes the focus to the growing world of AI, where security threats look very different from traditional bugs. The company says it wants experts to help identify “rogue actions,” those moments when AI behaves in unexpected or dangerous ways, like leaking personal data, executing wrong commands, or letting attackers manipulate connected devices.
What counts as an AI bug
Google has given a clear idea of what kind of vulnerabilities it wants researchers to look for. Imagine an attacker tricking Google Home into unlocking a smart door or using a hidden command that makes Gmail summarise someone’s emails and send them to a third party. These are the sort of high-risk exploits Google wants to uncover before they can be used in the real world.
However, the company has made one thing clear, simply making Gemini or any other AI model hallucinate doesn’t count as a bug. Problems related to the kind of content the AI generates, such as hate speech or copyrighted material, should be reported directly within the product using its feedback tools. Google says this helps its AI safety teams retrain and improve models in a more targeted way.
How much can you earn
The biggest rewards — up to $20,000 — are reserved for vulnerabilities found in Google’s key products like Search, Gemini apps, Gmail, and Drive. If a report stands out for its quality or originality, bonuses can push the payout up to $30,000. Lower but still meaningful rewards will be given for flaws discovered in other tools like NotebookLM or the experimental AI assistant Jules.
According to Google, this move is part of its wider effort to strengthen AI security as these systems become deeply integrated into its products. In fact, the company says researchers have already earned more than $430,000 in the past two years by exposing AI-related risks, even before this official program was launched.
Google launches CodeMender tool: AI fixing AI
Alongside the new reward program, Google also introduced a tool called CodeMender. It’s an AI agent designed to automatically detect and fix security vulnerabilities in open-source software. So far, CodeMender has helped patch over 70 verified issues after being checked by human reviewers. The company says tools like this show how AI can also play a positive role in making technology more secure.
