Google Uses AI to Fix Chrome Bugs

AI security is usually discussed as a threat: phishing, malware, model exploits and automated attacks. This story shows the other side — AI being used to repair software faster.

What happened

Google said internal AI tools helped fix 1,072 Chrome security bugs across two recent releases.

That is a unusually large number. Chrome is one of the most widely used browsers in the world, so security work happens at huge scale. Bugs in browser code can matter because browsers touch nearly everything users do online: passwords, sessions, payments, documents, workplace tools and personal accounts.

The interesting part is not just that AI helped write patches. It is that AI appears to be moving into the full software-security workflow: identifying risky code, assisting engineers, accelerating fixes and helping teams handle backlogs that would otherwise take much longer to clear manually.

Why it matters

For startups, this is a signal that AI-native security engineering is becoming a real category.

If large engineering teams can use AI to find and fix vulnerabilities faster, the same pattern will move into developer tools, code review, cloud security and application-security platforms. The winners will not simply be tools that say “AI for security.” They will be products that fit into how engineers already triage, verify and ship fixes.

There is also a competitive risk. If defenders can use AI to find bugs faster, attackers can too. That makes speed, evidence quality and automated remediation more important.

The bigger picture

Software security is becoming an automation race. The old model was mostly human experts reviewing code, filing tickets and patching issues. The new model will combine engineers, specialised models and agentic workflows.

That does not remove the need for human judgement. But it does change the economics: bug backlogs that once looked impossible may become manageable, while companies that cannot automate security work could fall behind.