Google Chrome to Adopt AI‑Driven Dynamic Patching and Twice‑Weekly Updates
Google is developing a “dynamic patching” system that can apply security updates to the Chrome browser without requiring a full restart, using the browser’s multi‑process architecture to replace background child processes on the fly. The effort is tied to a shift toward a two‑week release cycle for major Chrome milestones and a pilot of twice‑weekly updates, aiming to narrow the window between vulnerability discovery and user protection.
AI models such as Gemini have been used to discover and help fix vulnerabilities, including a critical sandbox‑escape bug (CVE‑2026‑3545) that had existed in Chrome’s code for 13 years. Google reports that 1,072 security flaws were fixed in Chrome versions 149 and 150, and Chrome 151 later patched 370 flaws, seven of them critical. These changes are intended to keep the browser continuously up‑to‑date and reduce exposure to fast‑moving AI‑powered attacks.
Entities: BigSleep · CVE‑2026‑3545 · Chrome 150 · Chrome 151 · Chrome web browser · CodeMender · Gemini · Gemini (Google AI model) · Gemini AI model · Google Chrome · Google LLC
Claims
What the coverage asserts, and how well corroborated each claim is across sources.
- [● 7 SOURCES] Google is developing dynamic patching that can apply Chrome updates without a full browser restart. (multiple)
- [● 6 SOURCES] Google is piloting a twice‑weekly update cadence for Chrome. (multiple)
- [● 7 SOURCES] A Gemini‑powered AI agent discovered a critical sandbox‑escape vulnerability (CVE‑2026‑3545) that had been present for 13 years. (multiple)
- [● 6 SOURCES] Google aims to shift to a two‑week release cycle for major Chrome milestones. (multiple)
- [● 2 SOURCES] Chrome 151 fixed 370 flaws, including seven rated critical. (single)
- [● 8 SOURCES] Google fixed 1,072 security flaws in Chrome versions 149 and 150. (multiple)
- [● 6 SOURCES] Dynamic patching leverages Chrome’s multi‑process architecture to replace background child processes on the fly. (multiple)