Credentials and secrets
Private keys, provider tokens, connection strings, passwords, assigned secret values and other supported credentials.
Google analytics are off.
Browser AI protection
Promptective protects typed, pasted, dropped, uploaded and submitted content on supported AI surfaces in managed browsers. Your organisation sets the rules for what can reach the provider and receives a clear record of every decision.
Browser entry channels
See policy follow typed, pasted, dropped and uploaded content before provider delivery.
Summarise the renewal notes for the account team.
Policy outcome
No findings. Ready for provider delivery.
Protection coverage
Promptective checks supported browser AI interactions for sensitive information and unsafe instructions, then applies your organisation’s rules before delivery.
Private keys, provider tokens, connection strings, passwords, assigned secret values and other supported credentials.
Email addresses, names, phone numbers, postal addresses, dates, government identifiers, account and payment details, health identifiers, private URLs and other supported sensitive information.
Instruction overrides, fake system boundaries, prompt extraction, role-play jailbreaks and forged tool structures in submitted content.
Browser compatibility
Deploy Promptective across Chrome, Microsoft Edge and Brave today, with Safari and Firefox support planned.
On supported AI surfaces, Promptective applies policy to typing, paste, drop, upload and submit. Blocked content stays available for the user to edit or remove.
Promptective maintains supported provider paths as interfaces change and records the active protection state so your team can verify where protection is active.
Deployment and evidence
Choose what your organisation allows. Promptective applies those rules across supported browser AI destinations and keeps a clear audit trail.
Set which supported AI destinations teams can use and how sensitive content is handled.
Warn users, remove supported sensitive values or stop unsafe content before it is sent.
See the AI destination, detected risk, decision and whether the content was sent.