anonym.legal

By · Last updated 2026-04-04

Atgal į BlogąAI Saugumas

DI Politika Be Technines Kontroles Nepavyksta

77% darbuotoju dalinasi jautriomis darbo duomenimis su DI priemonemis, nepaisant draudianciu politiku. Vyriausybes rangovas ikelejo FEMA potvynio paraisku duomenis.

April 4, 20268 min skaityti
AI data governancetechnical controlsChatGPT policy failureChrome Extension DLPenterprise AI security

Kai Politika Susiduria su Tikru Elgesiu

Vyriausybes rangovas buvo paspaustas. Jis turejo neapdorotu FEMA potvynio-lengvatos paraisku. Jis ikelejo vardus, adresus ir sveikatos irasus i ChatGPT, norydamas dirbt greiciau. Savo galvoje jis nelauke jokiu istatymu. Jis tiesiog naudojo geriausia priemone po ranka.

Rezultatas: vyriausybes tyrimas ir viesasis atskleidimas.

Tai yra pagrindines nepavykimas tik politika pagristu DI valdymo. Politika sako darbuotojams, ka daryti. Ji nesustabdo elgesio.

77% imones darbuotoju dalinasi jautriomis darbo duomenimis su DI priemonemis bent kartqui per savaigte - net kai politika tai draudia (eSecurity Planet/Cyberhaven 2025). Tai nera nerupestingi darbuotojai. Tai zmones, paspausti laiko, renkanties greicias priemones.

Kodel Politikos Subyra

DI naudojimo politikos remiasi zmogiskuoju spremu ivedumo taske. Tas momentas yra greitas. Darbuotojas gali neprisiminti politikos. Jie gali nevertinti turinio kaip "jautraus". Jie gali priimti rizika, nes laiko taupymas jaute labai didelis.

Cyberhaven 2025 m. 4 ketvirio analize nustate, kad 34,8% visu ChatGPT ivvestu duomenu turi konfidencialios verslo informacijos. Daugelis tu naudotoju zinojo politika. Jie vis tiek ikelejo.

Prieigos politikos veikia, nes sistemos jas priverstinai vykdo. DLP el. pasto sluoksnyje veikia, nes sistemos ji taiko. DI naudojimo politikos neturi priverstinio vykdymo ikelimo taske. Zmogaus sprendimas uzpildo ta spraga. Didesniu mastu zmones klysta.

FEMA rangovas padarydamas viena is tu klaidu. Jis nebuvo blogas veikejo. Priemone laimejo, nes politika jo praše rinktis letejimu uz greiti. Spaaudimo metu jis pasirinko greiti.

Technines Kontroles Sustabdo Tai, Ko Politika Negali

Vienintelis sprendimas, veikiantis didesniu mastu, veikia techniniame sluoksnyje - ne mokymo sluoksnyje.

Narskyles plekstinys gali perstverti mainines buferiu turini pries pat jis pasiekia bet kokia ziniatiinkline DI. Kai rangovas kopijuoja paraisku vardus ir adresus ir ikelia i ChatGPT, plekstinys aptinka AAS, anonimizuoja ji ir siuncia varia versija. DI mato [VARDAS_1] ir [ADRESAS_1] vietoj tikriuju verciu. Jis vis tiek atlieka uzduoti. Paraiskejos privatios detalies niekada nepasiekia ChatGPT serveriu.

Tai yra automatizuota. Ji nepraso naudotojo prisiminti nieko.

Kuretojams, naudojantiems Cursor arba GitHub Copilot, MCP Serveris teikia ta pati sluoksni. Kodas, ikeltas i DI konteksta, pirmiausia pereina per anonimizavimo varikliu. Kredencialai ir saviti identifikatoriai tampa zzetonais. DI gauna varia ivesti ir vis tiek pateikia naudingus rezultatus.

Ziurekite, kaip tai palygina su blokavamu: Blokavimas vs. Anonimizavimas - Narsykles DLP Palyginimas.

Kas Pasikeicia Su Techninems Kontrolemis

Su narsykles plekstiniu, FEMA rangovo scenarijus vykdomas kitaip:

  1. Rangovas kopijuoja paraisku irasus is bylu sistemos
  2. Plekstinys aptinka AAS spausklyje
  3. Perzuros modulas rodo, kas bus pakeista
  4. Anonimizuota versija eina i ChatGPT
  5. ChatGPT apdoroja uzklausa ir grazina rezultatus
  6. Rangovas gauna reikiama pagalba - tyrimas neprasideda

Politikos keisti nereikejo. Mokymas nevyko. Perstvertimo sluoksnis tai atsauge.

Politikos mokymas mazina rizika ribiniuose atvejuose. Technines kontroles pasalina nepaykimo rezi. FEMA incidentas buvo politikos nesekmes. Jis butu buvees neivykis, jei tas vienas Chrome plekstinys butu itaisytas to rangovo irenginyje.

Taip pat ziurekite:

Saltiniai

Susiję Straipsniai

AI Saugumas

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.

AI Saugumas

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.

AI Saugumas

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.

Pasiruošę apsaugoti savo duomenis?

Pradėkite anonimizuoti PII su 285+ subjektų tipais 48 kalbomis.

Pradėti Nemokamą Bandomąją VersijąPeržiūrėti Funkcijas

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.