Let me check is a free, privacy-first tool that tells you whether a question is safe to send to a public AI (ChatGPT, Claude, Mistral, Gemini) or whether it exposes personal or confidential data and should stay in your company's internal tools. Bilingual (English and French).
The analysis is 100% rule-based — no AI. Your text is matched against hundreds of words and phrases across categories such as financial data, personal identifiers, credentials, confidentiality mentions and documents. Each match adds a severity weight; the final score is 100 minus the total risk (100 = safe, 0 = high risk). Your questions are never stored — only anonymous statistics (score, categories, language) are kept.
Often not. Public AI tools may retain or reuse what you type. Personal data, payslips, bank details, passwords, API keys, client names or documents marked confidential should not be shared with a public AI. Let me check flags these risks before you send your prompt.
Paste your intended question into Let me check. It highlights sensitive segments, gives a 0–100 safety score, and tells you whether to use a public AI, keep it in your company's internal tool, or rephrase.
No. The database has no column to store the question text — only anonymous statistics (risk score, detected categories, language).
For teams and companies, a private deployment with custom detection rules is available. See Let me check for teams.