Google says that its internal artificial intelligence tools helped the company patch more security flaws in Chrome during June than it had corrected in the previous two years combined.
In a statement released Thursday, Google reported that it fixed 1,072 security bugs across the two most recent Chrome releases issued in June (Chrome 149 and Chrome 150). By comparison, the prior 23 releases over the last two years accounted for 1,036 fixes.
The company published a chart in a white paper detailing its efforts to apply AI to faster vulnerability discovery and remediation. Google refers to each release as a “milestone”; for context, the company notes that Chrome 126 was released in June 2024.
Doug Turner, Chrome’s director of engineering, told TechCrunch that large language models (LLMs) have “fundamentally shifted the economics of cybersecurity, transforming vulnerability discovery into an automated, industrial-scale operation.” Turner added, “By applying models like Gemini, we are preemptively fixing vulnerabilities, outpacing our adversaries and making Chrome safer with every update.”
Google is not alone in seeing this pattern. Earlier in the month, Microsoft said it had patched a record 570 security flaws across its product lines as part of its regular monthly updates (commonly known as “Patch Tuesday”), and attributed the jump in fixes in part to its use of AI.
Apple, however, does not appear to show the same exponential increase. An independent tally reports that Apple had fixed 482 bugs in 2026 so far, a figure roughly on track to match or exceed last year’s total and comparable to the number of bugs Apple fixed in 2015.
TechCrunch contacted Apple for comment but did not receive a response.
Why this matters
These figures align with long-standing warnings from cybersecurity experts that AI-driven tools would greatly increase the volume of detectable vulnerabilities, pressuring defenders to adopt AI-based techniques themselves. The specific comparison — 1,072 fixes in two June releases versus 1,036 fixes in the prior 23 releases over two years — illustrates the rapid change in scale.
The trend has implications for software development, patch management, and how companies assess and prioritize security risk going forward.
