Google Chrome CVE-2026-118331

Exploitation requires the victim to actively load an attacker-controlled crafted HTML page in Chrome (UI:R), so a drive-by or social-engineering lure is the delivery vector; no authentication is needed (PR:N). The affected component is the FileAPI, and the condition is a maliciously crafted input that defeats FileAPI's input validation. The key limiting factor is that resulting code execution is confined to the renderer sandbox (S:U, 'inside the sandbox') - it does not by itself grant host-level access, so real-world full compromise additionally depends on chaining a separate sandbox escape. Affects Chrome builds prior to 156.0.8078.73.

Signals are largely consistent and point to a genuine but not emergency-level priority. The CVSS 3.1 vector (AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H, 8.8) describes remote, low-complexity, unauthenticated exploitation that nonetheless requires user interaction - the victim must load a crafted HTML page. Chromium rates the security severity Critical, and the issue is tagged RCE, but the description explicitly scopes code execution to 'inside the sandbox' (consistent with S:U, no scope change), meaning this is renderer-level compromise rather than a full host takeover unless chained with a separate sandbox escape. The vuln.today decision label is 'this_week', not act_now/emergency, aligning with the absence of active-exploitation signals. No EPSS score, no CISA KEV listing, and no public exploit reference are present in the input, so there is no public exploit identified at time of analysis and no evidence of in-the-wild use. Net: a real, broadly-reaching browser flaw that warrants prompt patching on the vendor's weekly cadence, but the sandbox-confined impact and required user interaction temper it below drop-everything urgency.

A remote attacker hosts or compromises a web page containing crafted content that abuses improper input validation in Chrome's FileAPI. When a user with a vulnerable Chrome build (prior to 156.0.8078.73) visits the page, the flaw is triggered to achieve arbitrary code execution within the renderer sandbox. No public exploit code is identified in the available sources; full host compromise would require chaining this with a separate sandbox-escape vulnerability.