Like finding a side door in a building you already had to trick someone into letting you enter
CVE-2026-11026 is an *Extensions* policy-enforcement flaw in Google Chrome. Per Google and NVD, Chrome versions before 149.0.7827.53 on Linux and 149.0.7827.53/.54 on Windows and macOS let a crafted malicious extension bypass navigation restrictions after the attacker has already convinced a user to install that extension. The affected population is therefore Chrome endpoints running pre-fix builds *and* still allowing user-driven installation of an untrusted extension.
Google rated it Medium, and that is already on the generous side for enterprise prioritization. The decisive friction is not browser exposure; it is the requirement to land a malicious extension first, which is a separate social-engineering or policy-control failure and one many managed fleets materially reduce with allowlists, blocklists, forced installs, or permission gating.
4 steps from start to impact.
Land a malicious extension
CRX. The Chrome bug is not reachable until the extension is installed and enabled. The practical weapon here is the malicious extension itself, not a network exploit.- Target uses Google Chrome below
149.0.7827.53 - User can install extensions or local policy allows sideloading / store installs
- Attacker can socially engineer the user into installing the extension
- Managed Chrome fleets often use allowlists, blocklists, or forced-install-only models
- Users still see install flows and permission prompts unless admin policy suppresses them
- Security teams can detect new extension installs from browser management telemetry
Trigger the navigation-restriction bypass
- Malicious extension is installed and active
- Victim is running the vulnerable Chrome build
- Bug details are still restricted in Chromium, which slows commodity weaponization
- The flaw appears bounded to browser extension behavior, not a browser sandbox escape or OS compromise
Abuse the expanded browser action surface
- Extension has enough permissions and a user session worth abusing
- User browses to workflows the attacker wants to manipulate
- Impact stays inside the browser trust boundary
- No evidence in the public record of reliable one-click account takeover or host-level compromise from this bug alone
- Blast radius is per-user, per-browser profile
Deliver user-level integrity impact
- Targeted user remains active in the compromised browser profile
- The extension's bypass materially helps the attacker's workflow
- No KEV listing
- No public PoC found
- Very low EPSS supplied in the prompt
The supporting signals.
| In-the-wild status | No public evidence of active exploitation found in the sources reviewed, and it is not in CISA KEV. |
|---|---|
| Proof-of-concept availability | No public PoC or exploit repo located. The Chromium bug link currently resolves to a sign-in gate, and Google's release note says bug details may stay restricted until most users are patched. |
| EPSS | Prompt supplied 0.00008 (~0.008% probability over 30 days under the EPSS model). Percentile was not verified from FIRST in this review, but the score is floor-level low. |
| KEV status | Not listed in the CISA Known Exploited Vulnerabilities Catalog. |
| CVSS vector readout | CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:N/I:H/A:N means remote delivery with low complexity, but *user interaction is mandatory* and the impact is integrity-focused, not code execution. |
| Affected versions | Google/NVD describe Chrome prior to 149.0.7827.53 as affected; the stable release post specifically calls out 149.0.7827.53 for Linux and 149.0.7827.53/.54 for Windows and macOS. |
| Fixed versions | Patch baseline is 149.0.7827.53 on Linux and 149.0.7827.53/.54 on Windows/Mac per the Chrome stable update. |
| Exposure and scanning reality | This is a client-side browser bug, so Shodan/Censys-style internet exposure counts are largely meaningless. What matters is your internal Chrome version inventory and whether endpoints allow non-approved extensions. |
| Disclosure timeline | Reported by Google on 2026-03-30 in the Chrome release note; CVE published by NVD on 2026-06-04. |
| Reporter / researcher | Release note attributes the issue to Google rather than an external researcher. |
noisgate verdict.
The single biggest severity reducer is the attacker prerequisite: they must first get a user to install a malicious extension, which makes this a post-social-engineering control bypass rather than a clean remote exploit. In managed enterprises, extension governance sharply narrows the reachable population and keeps the blast radius at the user/browser-profile layer.
Why this verdict
- Major downward adjustment: malicious extension install is required. That prerequisite implies the attacker already succeeded at phishing, social engineering, or bypassing extension governance before this CVE matters.
- Enterprise exposure is narrower than Chrome's market share suggests. Fleets that enforce extension allowlists, blocklists, forced installs, or permission restrictions remove a large share of the reachable population from this attack path.
- Blast radius is local to the browser session. This is not pre-auth server compromise, wormable network exposure, sandbox escape, or host-level RCE; it is a browser-policy bypass that amplifies a bad extension foothold.
Why not higher?
There is no public evidence here of active exploitation, no KEV listing, and no public exploit chain that turns this into mass compromise. More importantly, the attacker does not get value until the victim installs a malicious extension, which is far more limiting than a normal drive-by browser bug.
Why not lower?
It is still a real Chrome security boundary failure in a ubiquitous product, and once a malicious extension is present this bug can widen what that extension can do. If your fleet permits broad user-driven extension installs, the residual risk is non-zero enough that it should still be patched in routine browser maintenance.
What to do — in priority order.
- Enforce extension allowlisting — Move managed Chrome browsers to approved-extension-only where operationally possible. For a
LOWverdict there is no SLA beyond backlog hygiene, but this is the highest-value control because it kills the main prerequisite rather than the bug symptom. - Block risky extension permissions — Use Chrome admin controls to deny extensions that request permissions your business does not need, especially broad browsing or data-access capabilities. For
LOW, treat this as hygiene work in the normal policy-tuning cycle rather than emergency mitigation. - Review newly installed extensions — Mine Chrome Enterprise or endpoint management telemetry for recently installed, sideloaded, or low-reputation extensions. This is useful immediately because the attack chain starts with extension installation, not network exploitation.
- Keep Chrome auto-update healthy — Browser auto-update is the cleanest remediation path for this class of client bug. For
LOW, fold straggler cleanup into standard browser-version compliance reporting rather than carving out a dedicated hotfix campaign.
- A WAF does not help; there is no server-side request path to filter.
- Perimeter internet exposure scanning does not help; this is a client-browser issue, not an exposed service.
- Generic network IDS signatures are weak here because the critical event is extension installation and in-browser policy bypass behavior.
Crowdsourced verification payload.
Run this on the target Windows endpoint or through your software-distribution / RMM tooling. Invoke it with powershell.exe -ExecutionPolicy Bypass -File .\Check-Chrome-CVE-2026-11026.ps1; standard user rights are usually enough because it only reads file and registry version data.
# Check-Chrome-CVE-2026-11026.ps1
# Purpose: Determine whether Google Chrome on a Windows host is vulnerable to CVE-2026-11026
# Logic: Vulnerable if any installed Chrome version is below 149.0.7827.53
# Exit codes: 0=PATCHED, 1=VULNERABLE, 2=UNKNOWN
$ErrorActionPreference = 'SilentlyContinue'
$minimumVersion = [version]'149.0.7827.53'
$foundVersions = New-Object System.Collections.Generic.List[version]
function Add-VersionIfValid {
param([string]$VersionString)
if ([string]::IsNullOrWhiteSpace($VersionString)) { return }
try {
$v = [version]$VersionString
$foundVersions.Add($v)
} catch {
# Ignore unparsable versions
}
}
# Common Chrome executable locations
$paths = @(
"$env:ProgramFiles\Google\Chrome\Application\chrome.exe",
"$env:ProgramFiles(x86)\Google\Chrome\Application\chrome.exe",
"$env:LocalAppData\Google\Chrome\Application\chrome.exe"
)
foreach ($p in $paths) {
if (Test-Path $p) {
try {
$file = Get-Item $p
Add-VersionIfValid -VersionString $file.VersionInfo.ProductVersion
Add-VersionIfValid -VersionString $file.VersionInfo.FileVersion
} catch {}
}
}
# Registry uninstall keys as fallback
$regPaths = @(
'HKLM:\SOFTWARE\Microsoft\Windows\CurrentVersion\Uninstall\*',
'HKLM:\SOFTWARE\WOW6432Node\Microsoft\Windows\CurrentVersion\Uninstall\*',
'HKCU:\SOFTWARE\Microsoft\Windows\CurrentVersion\Uninstall\*'
)
foreach ($rp in $regPaths) {
try {
Get-ItemProperty $rp | Where-Object {
$_.DisplayName -match 'Google Chrome'
} | ForEach-Object {
Add-VersionIfValid -VersionString $_.DisplayVersion
}
} catch {}
}
if ($foundVersions.Count -eq 0) {
Write-Output 'UNKNOWN'
exit 2
}
$uniqueVersions = $foundVersions | Sort-Object -Unique
$minFound = $uniqueVersions | Sort-Object | Select-Object -First 1
if ($minFound -lt $minimumVersion) {
Write-Output 'VULNERABLE'
exit 1
} else {
Write-Output 'PATCHED'
exit 0
}
If you remember one thing.
Sources
- NVD record for CVE-2026-11026
- Chrome stable channel update for Desktop, June 2 2026
- Chromium issue 497599683
- CISA Known Exploited Vulnerabilities Catalog
- FIRST EPSS model documentation
- Chrome Enterprise: Allow or block apps and extensions
- Chrome Enterprise policy: ExtensionInstallForcelist
- MITRE CWE-284 Improper Access Control
What defenders are saying.
Crowdsourced verification outputs.
Results submitted by users who ran the verification payload against their environment.