Skip to main content

Reporting a Security or AI Concern

Spot a possible security issue or an AI answer that looks wrong? Here is how to report it to security@metaintro.com, what to include and what happens next.

Enterprise

Paste it into Claude, ChatGPT, Codex, Cursor or Pi.

What to Report

  • A possible security issue, like a suspicious sign-in or a link that should not work.
  • Someone seeing information they should not.
  • An AI answer that is inaccurate, biased or harmful.

How to Report It

Email security@metaintro.com and include:

  • What happened and when
  • Screenshots, if you have them
  • The member or page involved. For anything sensitive, use their initials, not their full details.

What Happens Next

We look into every report. If we confirm a security incident involving your data, we notify your organization's designated contact within 72 hours and explain what happened, what data was involved and what we are doing about it.

When the AI Spots a Safety Concern

You do not have to catch everything yourself. When a member's message signals possible harm, Chat Safety pauses the chat, shows the member crisis support numbers, and emails their case manager, your admins and the Metaintro safety team. See Chat Safety: How Metaintro Keeps Conversations Safe.

If Someone May Be in Danger

Do not wait for us. Follow your organization's own safety procedures, and call 911 or the 988 Suicide & Crisis Lifeline.

Frequently Asked Questions

Will the member know I reported something?

No. Reports go to Metaintro, not to the member.

Should I report an AI answer that was just a little off?

Yes. Every report helps us improve.

Where do I send questions that are not about security?

Email support@metaintro.com. Security reports go to security@metaintro.com.

What to do next

Need Help?

Not sure whether to report something? Email security@metaintro.com and we will help.

Was this article helpful?

Return to navigation