By · Last updated 2026-03-30

Back to BlogAI Security

83% of AI Extensions Are Never Audited

83% of Chrome extensions with broad permissions have never been security-audited (USENIX 2025). 45% of enterprise employees use unapproved extensions.

March 30, 20268 minute read
Chrome extension security auditenterprise browser governanceAI extension riskunaudited extensionsDLP browser

Updated for 2026

Most AI Extensions Are Never Audited

The Chrome Web Store hosts over 180,000 browser extensions. Many — mainly AI tools — ask for wide access. They read every page you visit. They see your clipboard. They can block or alter network requests.

USENIX Security 2025 found that 83% of Chrome extensions with broad permissions have never been audited. Developers built them, published them, and users installed them by the millions. No one checked whether each tool does only what it claims.

That gap is structural. The Chrome Web Store scans for known malware. It checks policy rules. It cannot confirm whether data collection is fully disclosed. It cannot detect whether data flows to hidden third parties.

Half of Enterprise Workers Run Unapproved Tools

LayerX's 2025 Enterprise Browser Security Report found that 45% of enterprise employees use browser extensions that IT never approved. The pattern is common. An employee finds a useful tool. They install it. IT never finds out.

Combine 83% unaudited with 45% unapproved. Nearly half of enterprise employees may be running tools whose security has never been checked. Those employees handle sensitive company data every day.

For regulated sectors, the risk is direct. An HR employee using an unvetted tool that reads clipboard content may have sent personal data to an unknown third party. A lawyer using an unvetted AI writing tool may have sent client data to an unknown party. See our legal compliance guidance for how these risks map to GDPR, HIPAA, and related frameworks.

What the 900K-User Incident Shows

A reported incident in early 2026 shows the failure mode. Malicious Chrome extensions exposed the AI chat logs of an estimated 900,000 users. Around 600,000 came from one tool. About 300,000 came from another. Both appeared to offer real AI features. Both were listed in the Chrome Web Store. Both had large user bases.

Data theft was complete within 30 minutes of install. By the time researchers found the tools, nearly a million users had already lost control of their AI chat history. That included any sensitive content they had typed.

Incogni's 2025 study found that 67% of AI Chrome extensions collect user data. Collection practices, disclosure, and data targets vary widely across that group. See our security and compliance overview for how browser-level controls compare to trusting each tool's own conduct.

An Enterprise Governance Framework

Blocking all browser extensions is not realistic. The cost is too high. The right response is a governance framework that limits exposure to vetted, approved tools — mainly for AI use.

Extension allowlisting. Define which extensions are allowed on enterprise devices. Require a security review before adding any new tool. Use Chrome Enterprise policy to block installs outside the approved list.

Stricter review for AI tools. Any extension that handles AI prompts gets extra scrutiny. Check network traffic to see where data goes. Review the full permission scope. Verify publisher identity.

Browser-layer controls. For approved AI tools, apply controls that intercept sensitive content before it reaches AI providers. This removes the need to trust each extension's own conduct.

The 83% unaudited rate is not a problem users can solve. Users cannot audit Chrome extensions themselves. Enterprise governance — approved lists, policy enforcement, and tech controls — is the reliable answer. See our FAQ and glossary of browser DLP terms for more.

anonym.legal's Chrome extension runs PII scanning locally in the browser. No chat content reaches anonym.legal servers during scanning. The modified, masked prompt is what gets sent to the AI service.

Sources

Ready to protect your data?

Start anonymizing PII with 285+ entity types across 48 languages.

About this page

We update this page when our platform or the law changes.

Read our founder note for how we work.

Each change shows up in the timestamp at the top.

Related reading

We follow these rules

  • GDPR (EU 2016/679).
  • ISO/IEC 27001:2022.
  • NIS2 (EU 2022/2555).
  • HIPAA safe harbor under 45 CFR § 164.514(b)(2).

Our promise

We do not sell your data.

We do not train models on your text.

We store your files in Germany.

You can delete your account at any time.

You own your work.

Where we run

Our servers live in Falkenstein, Germany.

We use Hetzner. They hold ISO 27001 certification.

All data stays in the EU.

Backups run every day.

Need help?

Email support@anonym.legal.

We reply within one business day.

How we test

We run a full check suite on every release.

Each surface gets its own sweep script and report.

Human reviewers spot-check the output each week.

We track recall and precision on a labelled set.

Bad runs block the deploy.

What we never do

  • We never sell your information to third parties.
  • We never train models on what you upload.
  • We never keep your work after you delete it.
  • We never share keys with any outside firm.
  • We never run ads inside the product.

Plans in plain words

We sell credits, not seats.

One credit covers one short job.

Long jobs use a few credits each.

You can top up at any time.

Unused credits roll over each month.

Read the plans page for current rates.

Who built this

A small team of engineers and lawyers built this.

We ship from Europe and work in the open.

Our founder note spells out why we started.

Where to start

How the parts fit

A browser add-on cleans text inside Chrome.

A Word plug-in handles drafts in Office.

A small desktop tool works on whole folders.

An agent protocol link feeds large models safely.

All four share one core engine and one rule set.

Words from our team

We started this work after a lunch about cookies.

One friend kept getting odd ads on her phone.

We asked why a court file leaked through a draft.

We sketched the first build on a napkin that week.

By month three we had a tiny demo for a friend.

She used it on her first case the next day.

Common questions we hear

Can the tool read scanned PDFs? Yes, with OCR.

Does it work on long files? Yes, in small chunks.

Can I roll my own rule set? Yes, save it as a preset.

Does it run offline? The desktop build runs offline.

Do you keep my files? No, the cloud build wipes after each run.

Will it learn from my work? No, we never train on inputs.

A short tour of the workflow

Upload a file or paste a snippet of prose.

Pick the entities you want gone from the draft.

Choose a method: replace, mask, hash, encrypt, or redact.

Press run and watch the side panel show each hit.

Skim the result and tweak any rule that misfired.

Save the cleaned file or send it to a teammate.