Guard
Dashboard
← Back to warning

Report a warning

Help us improve AI tool safety by reporting risky or incorrect behavior.

Submitting anonymously

Sign in to track your report and attach verified evidence.

Incident context

0/2000
0/2000

Steps to reproduce

  1. 1

What happens next?

  1. 1. Our team will review your report.
  2. 2. We may follow up for more information.
  3. 3. If confirmed, we will improve protections and notify you.

Link evidence (optional)

Report summary

Category
Secret exfiltration
Impact
High
Applies to
OpenAI Codex
Visibility
0/2000

Your report helps make AI tools safer for everyone. We never share private reports without permission.

Cancel