Add DangerBot check for AI logging compliance

This commit introduces a new Dangerfile to check for AI logging compliance:

- Detect usage of GitLab::Llm::Logger.build.info and .error methods
- Warn when AI logging is not gated behind expanded_ai_logging flag
- List files with potentially non-compliant AI logging
- Add relevant labels to merge requests with AI logging

This check helps ensure adherence to GitLab's AI data usage policies and
promotes proper AI logging practices across the codebase.