Google’s Chrome Vulnerability Findings Surge, Possibly With AI’s Help
Summary
Google-attributed Chrome vulnerabilities rose to 100 in its May 5 advisory; more than 70 flaws patched in the two latest releases were found internally. AI’s role is suspected, not confirmed.
Key points
- Google-attributed vulnerabilities in Chrome advisories rose from a handful in late March and early April to 16 on April 15, 21 on April 28, and 100 on May 5.
- More than 70 vulnerabilities patched in Chrome’s two most recent releases were found internally by Google.
- Google has not confirmed that AI discovered the Chrome vulnerabilities or identified which model or tool may have been involved.
- Google said AI and automation are helping its teams find root causes, propose fixes, and identify variants of known problems.
- Google has developed AI-assisted vulnerability discovery tools including Big Sleep and CodeMender; Google describes CodeMender as able to identify vulnerabilities and recommend, test, and apply fixes with approval.
- Google did not respond to SecurityWeek’s questions about the number of vulnerabilities found using AI or the tools involved.
Article Details
- Event Type
- Google’s reported Chrome vulnerability discoveries surged, with AI suggested as a possible driver.
- Impact
- Google advisories listed 16 vulnerabilities reported by Google in the April 15 Chrome update, 21 in the April 28 update, and 100 in the May 5 advisory. More than 70 vulnerabilities patched in the two most recent Chrome releases were found internally by Google. Google has not confirmed that AI was used to discover them, nor disclosed the exact number found with AI or which model or tool may have been involved.
Vendors
AnthropicMozilla, for instance, found over 270 Firefox vulnerabilities with the aid of Anthropic’s new Claude Mythos model.GoogleThe number of Chrome vulnerabilities discovered by Google has surged over the past month, likely driven by the company’s use of AI.MicrosoftMicrosoft and Palo Alto Networks have also found many vulnerabilities in their products using advanced AI vulnerability discovery tools.MozillaMozilla, for instance, found over 270 Firefox vulnerabilities with the aid of Anthropic’s new Claude Mythos model.Palo Alto NetworksMicrosoft and Palo Alto Networks have also found many vulnerabilities in their products using advanced AI vulnerability discovery tools.
Products
Claude MythosMozilla, for instance, found over 270 Firefox vulnerabilities with the aid of Anthropic’s new Claude Mythos model.GeminiLeveraging Agent Platform capabilities and advanced Gemini models, CodeMender autonomously identifies vulnerabilities within your code,” Google said in a blog post this week.Google ChromeThe number of Chrome vulnerabilities discovered by Google has surged over the past month, likely driven by the company’s use of AI.Mozilla FirefoxMozilla, for instance, found over 270 Firefox vulnerabilities with the aid of Anthropic’s new Claude Mythos model.