Google Chrome CVE-2026-106228

Exploitation is gated by a hard precondition explicitly stated in the description: the attacker must FIRST have compromised the Chrome renderer process (e.g., via a separate renderer-stage bug) - this is not a standalone remote exploit. Given that foothold, the attacker then needs the victim to load a crafted HTML page (UI:R / user interaction required), and the specific affected component is the Google Lens feature acting as a confused deputy (CWE-441) to perform a privileged operation that crosses the sandbox boundary. Limiting factors: the required renderer compromise and user interaction, combined with high attack complexity (AC:H), mean this cannot be triggered against a default install without a prior exploitation stage; only Chrome versions prior to 155.0.8059.39 are affected, and updating resolves it.

This is a genuine high-value issue but not an immediate fire-drill, and the signals are internally consistent rather than conflicting. The CVSS 3.1 base score of 8.3 (AV:N/AC:H/PR:N/UI:R/S:C/C:H/I:H/A:H) is driven up by the scope change (S:C) inherent to a sandbox escape and by total CIA impact, but it is tempered by high attack complexity (AC:H) and required user interaction (UI:R). Critically, the description establishes that this is a second-stage vulnerability: the attacker must have ALREADY compromised the Chrome renderer process before this confused-deputy flaw (CWE-441) in Google Lens can be abused to execute code outside the sandbox. This aligns with Chromium's own 'Medium' security severity rating, which is lower than the 8.3 would suggest in isolation. The SSVC framework reinforces a measured posture: Exploitation=none (no observed activity), Automatable=no (cannot be mass-weaponized), but Technical Impact=total (full sandbox escape if chained). There is no CISA KEV listing and no public POC, so there is no public exploit identified at time of analysis. EPSS was not provided in the input and so cannot be weighed. Bottom line: prioritize patching because the affected product (Chrome) is ubiquitous and a vendor fix exists, but the practical barrier - needing a pre-existing renderer compromise plus user interaction plus high complexity - means this is a chain component for sophisticated actors, not a widely sprayable standalone exploit.

An attacker first achieves code execution inside Chrome's renderer process (typically by chaining a separate renderer RCE bug), then lures the victim to a crafted HTML page that abuses the Google Lens component as a confused deputy to perform privileged actions on the attacker's behalf, escaping the sandbox to run arbitrary code with browser-process privileges. No public exploit code or active exploitation has been identified for this specific CVE; exploitation requires the complex precondition of an already-compromised renderer, consistent with the CVSS AC:H and UI:R metrics.