Google says it fixed more Chrome bugs in June than over the past two years, thanks to AI
As experts have warned for the last two years, some companies — like Microsoft and now Google — are finding and patching an exponential number of bugs in their products, thanks to the use of LLMs and AI tools.
<p>Google's June Chrome patch count, reportedly the highest in two years, is being credited to AI-assisted vulnerability discovery. The framing matters. The volume of bugs fixed is not the same as the volume of bugs introduced; it is a function of how aggressively a codebase is being scanned.</p><p>Large language models are well suited to the mechanical parts of security work: pattern matching across millions of lines, generating fuzz inputs, triaging crash reports, and writing the boilerplate of a patch once a root cause is identified. That compresses the human bottleneck. A team that previously relied on a small number of senior engineers to review every report can now process a much wider funnel.</p><p>The implication for the labor market is narrow but real. Demand for traditional manual code auditors and junior security analysts is likely to soften, while compensation for engineers who can design, supervise, and validate AI-driven security pipelines should rise. The leverage shifts from finding bugs to defining what counts as a bug worth finding.</p><p>There is also a quieter signal here. When a vendor publishes a record patch number, it is simultaneously admitting a record number of latent defects. AI is not making software more secure in an absolute sense; it is making the discovery process faster and more thorough. The race between code generation and code review has tilted, for now, toward review.</p>