Google Chrome CVE-2026-106199

Exploitation requires the attacker to have ALREADY compromised the Chrome renderer process on Android (a prior, separate exploit) before this flaw can be used; it is a second-stage site-isolation bypass, not an initial-access bug. It additionally requires user interaction - the victim must load a crafted HTML page (UI:R). It is specific to Google Chrome on Android prior to 155.0.8059.39. The renderer-compromise prerequisite and required user interaction are the primary limiting factors, and they explain why both the vendor severity (Low) and EPSS (0.21%) are low despite the 8.2 CVSS score.

Signals conflict sharply here and must be weighed carefully. The CVSS 3.1 base score is 8.2 (High) with an AV:N/PR:N/UI:R/S:C vector implying a network-reachable, unauthenticated flaw with a scope change - but the vendor (Chromium) rates the security severity as 'Low,' and the description makes clear this is NOT a standalone remote bug: the attacker must ALREADY have compromised the renderer process before this site-isolation bypass is reachable. That prerequisite is the single most important risk-limiting factor and is not reflected in the PR:N metric (Chrome's scoring convention does not treat renderer compromise as a 'privilege'). EPSS is 0.21% (11th percentile), consistent with the vendor's Low rating and indicating very low near-term exploitation probability. The CVE is NOT on CISA KEV and no public exploit code is referenced. Net: despite the 8.2 headline, this is a second-stage sandbox/isolation-bypass primitive useful only to an attacker who already holds a renderer exploit - a meaningful component for a full chain but not a standalone priority. Treat as routine patch-cycle severity, not emergency, while noting that site-isolation bypasses are valuable building blocks in exploit chains. Discrepancy to verify: CVSS High vs vendor Low, and PR:N vs the description's explicit 'compromised the renderer process' precondition.

An attacker who has already exploited a separate renderer-process vulnerability (e.g., via a malicious ad or compromised site) lures a Chrome-on-Android user to a crafted HTML page. From within the compromised renderer, the attacker abuses the Actor authorization flaw to bypass site isolation and read data belonging to other origins. No public exploit is identified at time of analysis; the renderer-compromise prerequisite and required user interaction (visiting the page) substantially limit real-world use.