Privacy Shield

Names, emails, IBAN, addresses, medical IDs — removed from your question before any of the 6 AIs see it. Re-injected on your side after the answer comes back. The AI providers never see the originals.

Use Privacy Shield free

A first in multi-AI

Standard provider privacy modes promise one thing: your conversation will not be used for training. That is a useful promise, but it is a promise about what happens afteryour data has already left your device and hit the provider's servers in the clear.

Privacy Shield works one step earlier. Personal data is stripped from your question before any network call. The 6 AIs read an anonymized version. There is no original on the other side to leak, subpoena, or train on.

No other multi-AI product on the market today does pre-API PII stripping by default on the free tier. Satcove is the first.

How a question travels

Your question

Stays on your device with your real data.

PII stripped locally

Names, emails, IBAN, addresses replaced with placeholders before any network call.

6 AIs see only the anonymized version

Claude, ChatGPT, Gemini, Mistral, Perplexity, Grok — none of them see the real values.

Real data re-injected on your side

Satcove maps placeholders back to your real values in the verdict you read.

What gets stripped

Full names

First names, last names, initials in formal contexts.

Email addresses

Personal and professional, including aliases.

Phone numbers

International formats, EU, US, mobile, landline.

Postal addresses

Street, city, postal code combinations.

Bank details

IBAN, BIC, card numbers, account numbers.

Government IDs

Social security, national ID, passport, tax IDs.

Medical identifiers

Patient IDs, lab reference numbers, dates of birth in clinical context.

Credentials

API keys, tokens, passwords that slip into a paste.

Three levels of stripping

Level 1 — Standard

Direct identifiers stripped: full names, emails, phones, addresses. Sufficient for everyday questions involving people or documents.

Level 2 — Strong

Adds financial and government IDs: IBAN, BIC, social security numbers, tax IDs, passport numbers. Required for legal and finance use cases.

Level 3 — Clinical

Adds medical identifiers and contextual quasi-identifiers (dates of birth, rare conditions, hospital references). Required for lab result analysis and any clinical question.

Who it's for

Lawyers & in-house counsel

Ask the 6 AIs about a contract clause, a dispute, or a regulatory question without sending client identifiers to OpenAI, Anthropic, Google, or xAI.

Doctors & nurses

Get a multi-AI second opinion on a lab result or symptom set without leaking patient identity outside your jurisdiction.

Accountants & advisors

Run a tax or compliance question across 6 AIs without exposing IBANs, tax IDs, or company financials.

Anyone with a private question

Your medical history, your finances, your relationship, your job offer — your data, not OpenAI's training set.

Infrastructure

  • EU-region hosting on Vercel and Supabase, with automatic failover.
  • AES-256 encryption at rest, TLS 1.3 in transit.
  • Passwordless authentication. No password stored, none to leak.
  • Conversation content not stored on Satcove servers by default.
  • Your data is never used to train any AI model — provider or Satcove.

See /security for the full security posture and /gdpr for our GDPR commitments.

Frequently asked questions

What does Privacy Shield actually do?

Privacy Shield strips personal data from your question before any AI provider sees it. Names, emails, phone numbers, addresses, IBAN, government IDs, and medical identifiers are replaced with neutral placeholders client-side. The 6 AIs only ever see the anonymized version. When the verdict comes back, the placeholders are mapped back to your real values locally so you read a normal, complete answer.

How is Privacy Shield different from ChatGPT's privacy mode?

Standard provider privacy modes promise that your conversation will not be used for training. Your data still leaves your device, still hits the provider's servers, and is still readable by them in the clear. Privacy Shield removes the data before it leaves your device. There is nothing on the other side to leak.

Is Privacy Shield available on the free plan?

Yes. Privacy Shield is available on all Satcove plans, including Free. Privacy is not a paywalled feature — it is the default posture.

What types of personal data are detected?

Full names, email addresses, phone numbers, postal addresses, IBAN and bank details, government IDs (social security, tax, passport), API keys and tokens, and medical identifiers including patient IDs and clinically-sensitive dates. Three levels of stripping are available depending on your use case.

Is Satcove GDPR compliant?

Yes. Satcove is hosted in the EU, encrypts data at rest with AES-256 and in transit with TLS 1.3, runs on EU-region infrastructure (Vercel and Supabase EU), and does not store the content of your conversations on Satcove servers by default. Privacy Shield further reduces what AI providers see, narrowing your exposure surface.

Does any competitor offer pre-API PII stripping for multi-AI queries?

No. Provider privacy modes promise non-training, not non-exposure. Enterprise wrappers exist but require IT deployment and operate on single-model traffic. Satcove is the first consumer multi-AI product that strips personal data before any of the 6 model APIs see the request.

Ask your hardest, most private question. The AIs will never see who you are.

Available on every plan, including Free.

Satcove — A product by Abyssal Group