Use any AI.
Keep your secrets.

Tightlip swaps names, account numbers and patient details for placeholders on your device. The AI answers the placeholder version. Your device puts the real values back.

Run a free leak check Get USD 20 free

1. Your textOn your device

Privileged and confidential. Memo to file. Matter: Delgado v. Brennan Freight LLC, Case No. 1:24-cv-03187. Our client Maria Delgado (DOB 03/14/1981) says Brennan Freight LLC held back her final commission of $48,200 after she resigned. Opposing counsel is Daniel Okafor (dokafor@okaforlegal.com, 212-555-0143). Ms. Delgado wants any settlement wired to her account at Harbor Point Bank, routing 021000021, account number 4471009233.

2. What the AI seesSent to the AI

Privileged and confidential. Memo to file. Matter: [PERSON_1_LAST] v. [ORG_1], Case No. [CASE_NUMBER_1]. Our client [PERSON_1] (DOB [DOB_1]) says [ORG_1] held back her final commission of $48,200 after she resigned. Opposing counsel is [PERSON_2] ([EMAIL_1], [PHONE_1]). Ms. [PERSON_1_LAST] wants any settlement wired to her account at [ORG_2], routing [ROUTING_NUMBER_1], account number [BANK_ACCOUNT_1].

3. The answer you readRestored on your device

Asked: Draft a short demand letter to opposing counsel.Dear Daniel Okafor, We act for Maria Delgado in Delgado v. Brennan Freight LLC, 1:24-cv-03187. Your client held back her final commission of $48,200. Please confirm within 14 days that Brennan Freight LLC will pay it in full.

Real output from Tightlip on a made-up memo to file: 10 items protected (2 names, 2 organisations, 1 case number, 1 date of birth, 1 email address and 3 more). The swap table that links each placeholder to its real value stayed on the device.

State bars are telling lawyers to anonymize what they give AI.

Read two more opinions on the law page

“If client confidences and secrets are not protected by the GAI tool, then the lawyer must anonymize their inputs to protect client details.”

Alaska Bar Association, Ethics Opinion 2025-1, 23 April 2025

Tightlip does the anonymizing step on your computer, before the prompt goes out.

Bar opinions are guidance. Your jurisdiction’s rules and your client’s instructions govern.

Three steps, all on your side of the wire.

1
Find
On your device, Tightlip reads what you are about to send and finds names, account numbers, case numbers, patient IDs, keys and more.
2
Swap
Each value becomes a placeholder such as [PERSON_1]. The same person always gets the same placeholder, so the AI can still follow who did what.
3
Restore
The AI answers with placeholders. Tightlip puts the real values back on your device. The swap table never leaves it.

Exactly what leaves your device.

Your security team will ask. Here is the whole answer. The full data flow, with every permission the extension asks for, is on the security page.

Read the data flow

To the AI provider: The protected text, with placeholders such as [PERSON_1] in place of real values.
To Tightlip: A count of credits used. In the web workspace, the protected text also passes through our server on its way to the model. We do not store it or log it.
Stays with you: Your original text and files stay on your device. So do your always-protect list and the swap table that turns placeholders back into real values.

Built for people with a duty to keep quiet.

Use the AI you already have.

Bring your own ChatGPT, Claude or API key, or use ours in the web workspace. The protection works the same way in all four.

Web workspace

For anyone. Nothing to install.

Paste text or drop a PDF or Word file. See exactly what the AI will get, then chat. The AI model is included in the credit.

Browser extension

For people who already use ChatGPT, Claude, Gemini, Copilot or Perplexity.

Protects the message box when you press send and restores the reply on the page. You keep your own AI account.

SDK

For teams shipping AI features.

Wrap your OpenAI or Anthropic client in two lines. Personal data it finds is swapped before the request leaves your server.

Local proxy

For tools you cannot change.

Run one command and point any OpenAI- or Anthropic-compatible tool at it. Requests are protected and replies restored on your machine.

USD 0.05

per protected prompt. No seats, no subscription.

USD 20 of free credits when you sign up. That is 400 protected prompts.
Top up from USD 100 when you need more.
With the extension, SDK and proxy you only pay when Tightlip actually protects something.

See pricing

Where Tightlip stops.

Detection can miss things. Tightlip shows you what the AI will see before you send it. Add client names, matter names and project code names to your always-protect list.

Your AI policy still applies. Tightlip reduces what reaches an AI tool. Use it with AI tools your organisation already allows.

The answer is still the AI’s answer. Read it with the same care you would give a junior colleague’s draft.

Paste something real. See what would have leaked.

Run the free leak check

Use any AI.Keep your secrets.