Data loss prevention for generative AI. Brightrail detects sensitive information and helps stop it from leaving your organization.
Users
1
store count
Rating
0.0 / 5
unchanged in 30 days
Ratings
0
total ratings
Store rank
#365K
#299K among extensions
Version
0.18.0
updated 19 Nov 2026
Publisher and listing changes
From our daily record of this listing since 19 Nov 2026
No change of publisher name, contact email, website, privacy policy or listing name since 19 Nov 2026.
About Brightrail
Brightrail stops employees from sending sensitive data to generative AI assistants by mistake.
When someone sends a prompt or attaches a file in ChatGPT, Claude or Gemini, Brightrail checks it on their computer first. If it finds sensitive information, it pauses the send and explains what it found. The employee can redact the data and send, edit the message, or cancel.
WHAT IT DETECTS
• Payment card numbers
• Social Security numbers and ITINs
• Bank account numbers, IBANs and routing numbers
• API keys, access tokens and private keys
• Passwords and database connection passwords
• Passport numbers
• In attached files: source code and documents marked confidential
HOW IT WORKS
• Checks run on the computer. Detection uses built-in rules and a small AI model that ships inside the extension. Prompts and files are never sent anywhere to be checked.
• "Redact it and send" replaces each value with a placeholder, such as ‹Card ending 8812›, and sends the safe version.
• Attached files are checked when they're attached, before the assistant uploads them. Common document, spreadsheet and code files are supported.
• Works only on ChatGPT, Claude and Gemini. It does nothing on any other website.
BUILT FOR ORGANIZATIONS
Brightrail is installed for a whole company by its Google Workspace admin, from the Brightrail console. Admins see which employees are covered and which data types were stopped, so they can coach people and manage AI risk. Employees see a short "Brightrail is on" note the first time, and a toolbar icon that shows whether protection is active.
PRIVACY
• The original prompt, the sensitive values and file contents never leave the computer.
• When something is stopped, the organization receives a record with the time, the AI tool, whether the account was work or personal, the kinds of data found and what the employee did next.
• Only when the employee chooses "Redact it and send", the organization also receives the redacted message, exactly as the AI tool received it.
• Once a day, the extension tells the organization it's installed, using the work email signed in to Chrome.
• No ads, no data sales, and no training on customer data.
Installed without an organization, Brightrail still checks prompts on the computer, but the toolbar shows "Not set up by your organization" and nothing is reported.
Privacy policy: https://brightrail.com/privacy-policy
Version history
1 versions recorded
No update since 19 Nov 2026 changed the permissions or site access.