Daily briefingRead in 1 or 5 minutesUpdated ...

Live wire

TechCrunch · ...Friend, the lonely AI wearable, returns with a new voice and a much bigger price tag Gizmodo · ...Can Winamp Ride the Retro Tech Wave All the Way Onto Spotify’s Private Beach?Gizmodo · ...Casio’s Tiny Ring Watch Is Now a Full-Blown Oura Alternative Gizmodo · ...Scientists Find Evidence of Lost Amazonian Civilization That May Have Been Home to Millions CNET · ...Winamp Plans Comeback With New Deezer Partnership View all headlines

Latest editorial

Updated ...

Google patched over 1,000 Chrome bugs in two releases using AI — more than the previous 23 milestones combined

Google says large language models now handle nearly every stage of Chrome's vulnerability management pipeline, from discovery to patch generation.

Key points

Google patched 1,072 security bugs across Chrome 149 and 150, exceeding the total fixed in the previous 23 Chrome milestones combined.
LLMs are now used across Chrome's entire vulnerability management pipeline: discovery, reproduction, severity assessment, developer assignment, patch generation, and test creation.
Google's AI security tooling evolved from LLM-assisted fuzzing (2023) to Project Zero's Naptime, then Big Sleep with DeepMind, and a Gemini-powered agent harness in early 2026.

Security advisory

CVE:: CVE-2026-15903
Severity:: High
Affected:: Google Chrome desktop versions prior to 150.0.7871.128; Chromium-based browsers embedding the affected V8 engine; all operating systems (Windows, macOS, Linux) shipping vulnerable Chrome versions
Patch status:: Patched in Chrome 150.0.7871.128

3 sources 2 web 88% confidence

Source highlights

TechCrunchPrimaryGoogle says it fixed more Chrome bugs in June than over the past two years, thanks to AI—80%

bleepingcomputer.comWeb contextGoogle says AI helped Chrome fix 1,072 security bugs in two releases—64%

sentinelone.comWeb contextCVE-2026-15903: Google Chrome V8 RCE Vulnerability—48%

Full story and ongoing updatesRead full story

Story brief

Google says large language models now handle nearly every stage of Chrome's vulnerability management pipeline, from discovery to patch generation.

Who should care

DevelopersAudit your own CI/CD pipeline for opportunities to integrate LLM-assisted fuzzing, triage, and patch-generation tooling similar to Google's approach.

Google's results show AI can scale vulnerability discovery and remediation across a massive codebase; smaller teams may see proportional gains.

FoundersEvaluate whether AI-assisted security tooling can be a differentiator or cost-saver for your product, especially if you ship a large codebase or browser-based software.

The gap between AI-augmented and manual security workflows is widening rapidly; early adoption could reduce both breach risk and engineering cost.

PMsTrack the ratio of AI-discovered vs. manually reported bugs in your product and set targets for increasing AI-assisted coverage over the next two quarters.

Google's 1,072-bug milestone demonstrates that measurable AI-assisted security metrics are becoming a credible signal of product maturity.

What changed

2023Google began using LLMs to improve security fuzzing for Chrome.
Early 2026Google created a Gemini-powered agent harness to search the broader Chrome codebase for vulnerabilities while reducing false positives.
2026-07-23CVE-2026-15903, a high-severity V8 RCE vulnerability in Chrome, was published.