AI data security
Use any AI.Keep every secret.
Prompt freely. Leak nothing. Sentlyx finds sensitive data in every prompt and masks it on the laptop, before ChatGPT, Claude or any other AI app can see it.
SentlyxCoverage
Works where your teams use AI
Sentlyx covers the AI apps that your employees use, in the browsers and on the systems that you already manage.
AI apps
- ChatGPT
- Microsoft 365 Copilot
- Gemini
- Claude
- Microsoft Copilot
- Perplexity
- DeepSeek
- Grok
- GitHub Copilot
- Meta AI
- Qwen
- Kimi
- Mistral Le Chat
Browsers
- Google Chrome
- Microsoft Edge
- Firefox
- Brave
Operating systems
- Windows
- LinuxNext
- macOSNext
By default, raw values never leave the laptop. If you turn on full-prompt capture, the laptop encrypts the prompt with your key before it sends it.
Read the security overviewFAQ
Questions from security teams
Short answers here. The security overview has the detail.
Does Sentlyx see our prompts?
By default, raw values never leave the laptop. The Sentlyx agent masks sensitive values on the laptop, and Sentlyx gets the masked text and the metadata.
Full-prompt capture is off by default. If you turn it on, the laptop encrypts the prompt with your key, and only people with the Investigator role can read it. Your data stays in its home region. Today that is India; regions in Europe and the Americas are planned.
How do we roll it out?
Push the Sentlyx agent to every Windows laptop with Microsoft Intune or Configuration Manager. The install is silent. Linux and macOS come next. Admins sign in with SAML or OIDC single sign-on, SCIM keeps the users in sync, and the events go to Splunk, Microsoft Sentinel, IBM QRadar or a webhook.
What if an employee stops the agent or uses another browser?
The agent works at the network level of the laptop, so another browser does not get around it. You get an alert when an agent stops or when someone turns off a part of it. Remove local admin rights with your MDM, so that users cannot stop it. Sentlyx cannot see a personal phone or a personal laptop.
What does Sentlyx find?
Credentials such as API keys, cloud keys and passwords. Personal data such as email addresses, phone numbers, Aadhaar, PAN, SSN and card numbers. Code that matches your repositories. Your own terms, such as project code names. Sentlyx checks file uploads too.
Does masking break the AI answer?
No. The AI app gets a placeholder such as [EMAIL_1] and answers as usual, with the same placeholder in its answer. With the optional browser companion, Sentlyx puts the real value back on the screen.
Will it block too much?
Start in monitor-only mode, and see what Sentlyx finds before it acts. Then select an action for each type of data: allow, warn, redact or block. With warn, the employee types a reason and sends the prompt.
Say yes to AI, safely.
Book a 30-minute demo and see Sentlyx on your own policies.