Google credits AI with 1,072 Chrome security fixes in two June releases

Google said on Thursday that Chrome 149 and 150, both released in June 2026, fixed 1,072 security bugs. The previous 23 milestones, spanning roughly two years, delivered 1,036 fixes in total.

AI is changing vulnerability economics

Large language models can inspect code and identify weaknesses at a speed that manual security teams cannot match. The resulting increase in findings raises the pressure on vendors to validate reports, develop patches and distribute updates before attackers exploit the same flaws.

Google’s figures suggest that this shift has moved beyond forecasts. A chart published with its white paper shows a sharp rise in Chrome fixes between milestone 126, released in June 2024, and the two latest releases. The same pattern is visible in AI findings in FFmpeg and 429 Chrome 149 patches, which connected automated discovery with a rapidly expanding patch workload.

Google and Microsoft report record volumes

Doug Turner, Chrome’s director of engineering, said Google is applying models such as Gemini to find and repair vulnerabilities preemptively. The company describes the process as automated discovery operating at industrial scale.

“LLMs have fundamentally shifted the economics of cybersecurity, transforming vulnerability discovery into an automated, industrial-scale operation. By applying models like Gemini, we are preemptively fixing vulnerabilities, outpacing our adversaries and making Chrome safer with every update.”

Microsoft reported a similar jump in July, patching a record 570 security flaws across its product lines during Patch Tuesday and also citing AI use. Apple has not shown the same acceleration: an independent count puts its 2026 total at 482 fixes, broadly on pace with last year and close to its 2015 figure.

For businesses, higher bug counts should not be mistaken for declining software quality alone. They increasingly reflect stronger detection. The practical response is to shorten browser update windows, monitor deployment across managed endpoints and ensure that patch operations can absorb AI-driven growth in vulnerability disclosures.