ai-safe2-framework

AI SAFE² Support Policy

Support channels

This open-source project does not promise service-level response or resolution times. Security reports are acknowledged under the timeline in SECURITY.md. Other reports are prioritized by severity, reproducibility, affected users, and maintainer capacity.

Compatibility and maintenance

What maintainers need

Include the affected version, operating system, installation method, command or integration surface, minimal reproduction, expected and observed behavior, and sanitized output. Remove credentials, prompts, customer data, and other sensitive material.

Maintainers may close reports that cannot be reproduced, concern unsupported versions, duplicate an existing report, or request individual compliance certification. AI SAFE² outputs are decision-support evidence; organizational conformance and risk acceptance remain with accountable owners.