Google Chrome CVE-2026-13856

Exploitation requires that the attacker has ALREADY compromised the Chrome renderer process on Android - this is not a standalone remote bug but the second stage of a chain. The specific trigger is a crafted HTML page interacting with the Speech component (Web Speech API surface); the CVSS vector's UI:R indicates user interaction is also needed, and AC:H reflects the non-trivial conditions to reliably win the validation gap. Limiting factors: it only affects the Android build of Chrome below 150.0.7871.47, it cannot be used against a target whose renderer has not first been compromised, and the Medium Chromium rating plus low EPSS reflect that these prerequisites meaningfully constrain real-world use. Note the input's PR:N is Google's renderer-relative scoring convention and understates the prior-compromise prerequisite.

Signals point to a real but conditional issue rather than a mass-exploitation emergency. The published CVSS 3.1 vector (AV:N/AC:H/PR:N/UI:R/S:U/C:H/I:H/A:H, score 7.5) is inconsistent with the description: it lists PR:N (no privileges) yet the text explicitly requires an attacker 'who had compromised the renderer process,' which is a substantial prerequisite - Google's convention scores renderer-relative bugs as PR:N, but in real deployment terms this needs a chained prior exploit, and AC:H plus UI:R already reflect the added difficulty. EPSS is low (0.21%, 11th percentile) and there is no KEV listing and no public exploit identified, so this is not under active exploitation. Chromium's own 'Medium' severity rating is consistent with a bug only reachable after a renderer is already owned. Net priority: patch on the normal Chrome update cadence; it is not a drop-everything item, but it is a valuable link in a full renderer-to-browser exploit chain and should not be deferred indefinitely.

An attacker first lands a separate renderer-compromise exploit (for example via a malicious ad or web page that exploits an unrelated Chrome renderer bug), then serves a crafted HTML page whose Speech-related interactions send malformed input across the renderer-browser boundary. Because the Speech component under-validates that input, the attacker escalates out of the renderer sandbox toward higher-privilege browser-process context on the Android device. No public exploit is identified at time of analysis, and the high attack complexity plus required user interaction make opportunistic use unlikely.