Chrome pushes twice-weekly security fixes after AI finds 1,072 bugs

Google Chrome released 1,072 security fixes in its two major version updates in June 2026 — more than the previous 23 releases combined — and has since begun issuing patches twice a week, the browser’s security team said in a report published July 30.

The spike follows an intensification of the team’s use of artificial intelligence to find, assess and repair bugs. Chrome has deployed machine learning for vulnerability discovery since at least 2012, but Vice President and General Manager Parisa Tabriz called 2026 an “inflection point” for both offense and defense in software security.

Doug Turner, Chrome’s director of engineering, said the volume of fixes made the accelerated schedule sensible, though he questioned whether the twice-weekly cadence would be permanent.

The AI models are trained on every past vulnerability and every line of Chromium’s code history, allowing the tools to detect weaknesses even in features that have fallen dormant, such as printing.

Tabriz and Turner also emphasized structural shifts, including rewriting portions of C++ code in the memory-safe Rust language to eliminate entire classes of bugs.

She expressed hope that broader AI integration will make software more secure, but cautioned that improvements “will not come without effort.”