Thanks to its internal AI tool, Google said it discovered and fixed more security holes in the Chrome browser in one month than the previous two years combined.
Google announced that it successfully patched 1,072 security bugs on the two latest versions of Chrome, including Chrome 149 and 150, released together in June. This number surpasses the total of 1,036 vulnerabilities. fixed on 23 versions of Chrome over the past two years.
Information mentioned in Google’s report about using AI to proactively find and handle security weaknesses with outstanding speed. Since the LLM large language model exploded, cybersecurity experts predict that AI will help detect a large number of vulnerabilities, forcing defense units to apply AI to stay one step ahead of hackers.
Google Chrome logo at Google I/O, held in California, USA, May 2026. Photo: Luu Quy
Doug Turner, Chief Technical Officer of Chrome, shared: “The large language model has fundamentally changed the economics of cybersecurity, turning error detection into an automated process on an industrial scale.”
He added, by applying models like Gemini, the company is proactively fixing weaknesses before bad guys can exploit them, making Chrome safer with each update. update.
According to TechCrunch, Google is not the only unit that has recorded an explosion in the number of errors resolved using AI. In early July, Microsoft also announced a record number of patching 570 security vulnerabilities across all product lines in the “Patch Tuesday” periodic update. The company admitted that the sudden increase came from introducing AI into the system review process.
Unlike Google and Microsoft, Apple has not recorded similar growth. Independent statistics show that Apple fixes 482 bugs in 2026, equivalent or slightly higher than last year, and on par with the number of bugs the company handled in 2015.