Google исправила в Chrome за июнь 2026 больше багов, чем за два года — благодаря ИИ

In June 2026, Google announced an unprecedented leap in Chrome browser security work — in a single month, the company found and fixed more vulnerabilities than in the previous two years combined. The reason was the widespread use of large language models (LLM) and specialized AI tools for automated code bug detection.

How AI Changed the Speed of Bug Fixing

Google began systematically applying LLM to analyze Chrome's codebase — and the result was unprecedented. June 2026 set a record: the number of bugs found and fixed exceeded the combined total for the entire preceding two-year period.

Chrome is one of the world's largest browsers, with a codebase of tens of millions of lines in C++ and other languages. Traditional security audits at such code volumes required hundreds of engineer-hours and covered only a limited fraction of the codebase. Language models are radically changing this equation: LLM can analyze enormous volumes of code in search of known vulnerability classes — buffer overflows, use-after-free, unsafe memory operations, injection vectors — far faster and more systematically than any manual review.

In addition, LLM can detect context-dependent bugs: those that only manifest under a certain sequence of calls or specific conditions. Traditional static analyzers on such tasks often produce false positives or miss the threat entirely.

• June 2026 — a record month for the number of Chrome vulnerabilities fixed
• The result of one month exceeded the combined total for the previous two years
• Tools: large language models (LLM) and AI systems for static code analysis
• Microsoft recorded a similar effect for its own products

Why This Does Not Surprise Experts

Cybersecurity specialists had been warning of this shift approaching for the past two years: LLM can fundamentally change the balance of power between attack and defense in software.

"Companies like

Microsoft and now Google are finding and fixing exponentially more bugs in their products thanks to the use of LLM and AI tools," — according to TechCrunch.

Microsoft — a technology giant with a comparable codebase — also publicly confirms a sharp increase in the number of detected and fixed vulnerabilities following the adoption of AI tools. According to TechCrunch, the trend appeared simultaneously at two of the largest technology vendors — a sign of a systemic industry shift, not an isolated success of a single team.

Importantly, the acceleration is two-sided: the same LLM that help Google find and fix bugs could theoretically allow malicious actors to discover vulnerabilities faster. The balance between the growing number of fixes and the possible increase in attacker activity remains a key question for the industry in the coming years.

What This Means

The use of LLM for vulnerability detection is moving beyond experimentation and becoming standard engineering practice at the largest technology companies. If Google receives more fixes in one month than in the previous two years, this represents a structural turning point for the entire cybersecurity industry: the level of protection of mass-market software will grow significantly faster than before. The question is not whether AI will change security work — it already is.