anonym.legal

By · Last updated 2026-04-16

Nazaj na blogVarnost umetne inteligence

Po incidentu z razsiritvijo pri 900.000 uporabnikih

Januarja 2026 sta dve zlobni razsiritve za Chrome, ki sta jih namestilo vec kot 900.000 uporabnikov, vsakih 30 minut eksfiltrirali celotne pogovore v ChatGPT in DeepSeek.

April 16, 20268 min branja
malicious Chrome extensionAI extension security auditextension trust verificationlocal processing architecture900K extension incident

Incident januarja 2026

Posodobljeno za leto 2026. Januarja 2026 sta bili odkriti dve zlobni razsiritve za Chrome z vec kot 900.000 uporabniki.

Njuni imeni sta izgledali kot pravi orodji za umetno inteligenco:

  • "Chat GPT for Chrome with GPT-5, Claude Sonnet and DeepSeek AI" - vec kot 600.000 uporabnikov
  • "AI Sidebar with Deepseek, ChatGPT, Claude and more" - vec kot 300.000 uporabnikov

Obe sta poceli enako. Vsakih 30 minut sta poslali celotne pogovore z umetno inteligenco na oddaljeni streznik. Ukradeni podatki so vkljucevali kodo, osebne podatke, pravne zapiske in poslovne nacte. To je potrdil Astrix Security.

Ti razsiritve so zahtevali dovoljenje za "zbiranje anonimnih, neidentifikabilnih analiticnih podatkov." Ta besedna zveza se zdi varna. Ni bila. Zbrani podatki so bili popolnoma identifikabilni in zelo obcutljivi.

Tezava varnostne inverzije

Uporabniki, ki namestijo orodja za zasebnost z umetno inteligenco, zelijo zascito. Primer iz januarja 2026 kaze najslabsi mozni izid: orodje, ki ste ga namestili za zasebnost, je tisto, ki krade vase podatke.

To ni teorija. Dogajalo se je 900.000 uporabnikom hkrati. Pregledovanje v Googlovi spletni trgovini Chrome tega ni zaznalo. Uporabniske ocene tega niso razkrile. Kraja je bila skrita kot "analitika".

Incogni je ugotovil, da 67 % razsiritev za Chrome z umetno inteligenco aktivno zbira podatke uporabnikov. Za IT ekipe kljucno vprasanje ni: "ali to zbira kakrsne koli podatke?" Vprasanje je: "ali lahko preverim, da ta razsiritev ne more poslati vsebine pogovorov tretji osebi?"

Test preverjanja arhitekture

Obstaja ena zanesljiva kontrola za lokalno obdelavo: nadzor omrezja.

Razsiritev, ki lokalno zazna osebne podatke (PII), ustvari nic odhodnega prometa med zaznavanjem. Med korisnikovim lepljenjem in oddajo na platformo umetne inteligence se ne sme prikazati nobena povezava z zunanjim streznikov. Gre ven samo obdelani poziv.

Razsiritev, ki usmerja promet prek posrednika, poslje vaso vsebino na streznik tretje osebe. Ta operater streznika je zdaj v vasem modelu groznje.

Koraki IT preverjanja so preprosti:

  1. Namestite razsiritev v nadzorovano omrezje
  2. Zazenite testne pozive
  3. Preverite odhodne povezave s strezniki zaloznika med obdelavo PII

Ce ne prestane tega testa, je ne odobrite. Trzne izjave niso pomembne. Omreznemu prometu lahko zaupate. Glejte, kako to obravnava anonym.legal v nasem pregledu varnosti razsiritve za Chrome in vodniku za skladnost.

Kaj morajo zahtevati IT ekipe

Po januarju 2026 mora biti meja za orodja AI v brskalniku visja.

Minimalni seznam:

  • Lokalna obdelava - preverjena z omrezno revizijo, ne le trdena
  • Znan zaloznik - resnicno podjetje, jasen poslovni model
  • Neodvisna certifikacija - ISO 27001 ali enakovredno
  • Brez usmerjanja prek streznikov razvijalca za osnovne funkcije zasebnosti

Vecina orodij AI za brskalnik tega seznama ne bo prestala. 67-odstotna stopnja zbiranja podatkov to jasno kaze. Visoko stevilo namestitev ni signal varnosti. Orodja iz januarja 2026 so imela stotisoc uporabnikov, preden jih je kdorkoli preveril.

Za vec o varnih orodjih AI za brskalnik glejte nas pregled varnosti in skladnosti.

Viri

Sorodni članki

Varnost umetne inteligence

AI Coding Assistants Leak Production PII

Unit test fixtures with real customer records. Log files with production data for debugging. GitHub found 39 million secrets leaked in 2024.

Varnost umetne inteligence

Internal Wiki PII: Confluence Customer Data

Support teams document processes with screenshots of customer accounts. Over 3 years, that's thousands of GDPR data minimization violations in your.

Varnost umetne inteligence

Screenshot PII: Leaks in Internal Tools

Slack, Teams, Jira, and email regularly receive screenshots containing customer PII. This access-control violation bypasses every DLP tool.

Ste pripravljeni zaščititi svoje podatke?

Začnite z anonimizacijo PII z več kot 285 tipi entitet v 48 jezikih.

Začnite brezplačno preizkušnjoOgled funkcij

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.