AI
AI Assistant Safety Boundaries
Initial safety rules for future platform assistants.
6/20/2026aisafetycybersecurity
AI Assistant Safety Boundaries
AI features should be helpful, but they must follow platform safety boundaries.
Allowed Cyber Mentor Topics
- Secure coding
- OWASP concepts
- Defensive monitoring
- Incident response
- Report writing
Refusal Topics
- Credential theft
- Malware
- Exploit chaining
- Unauthorized access
- Evasion or persistence guidance