Executive brief
A vulnerability in Google Chrome's PrivateAI component could allow an attacker to escape the browser's security sandbox. To exploit this, an attacker must trick a user into visiting a malicious website and performing specific interactions or gestures. A successful attack could lead to full system compromise by allowing malicious code to run outside the restricted browser environment.
Technical details
A use-after-free (UAF) vulnerability exists in the PrivateAI component of Google Chrome. The flaw is triggered when a remote attacker convinces a user to engage in specific UI gestures while visiting a specially crafted HTML page. This memory corruption issue can be leveraged to bypass the Chrome sandbox (Sandbox Escape), allowing for arbitrary code execution with the privileges of the user. The vulnerability was addressed in Chrome version 147.0.7727.55. While Chromium developers classified the severity as Medium, CISA-ADP assigned a CVSS score of 9.6 (Critical) due to the potential for a full sandbox escape.
Affected products
- Google Chrome prior to 147.0.7727.55
Timeline
- 2026-02-18: other: Vulnerability reported to Chrome by Krace
- 2026-04-07: patched: Chrome 147.0.7727.55 released to stable channel
- 2026-04-08: disclosed: Initial CVE publication