Privacy before convenience
Personal data, secrets, credentials and proprietary code must not be sent to external tools without explicit authorisation.
lagrotti.dev
Responsible AI use
Boundaries, privacy, human review and transparency when using AI tools.
AI as an assistant, not the decision maker
Suggestions are checked against context, tests, sources and human review before entering production.
Transparency in use
When AI participates in content or code creation, that use should be reviewed and disclosed according to the material's context.
Security across the workflow
Model outputs are treated as untrusted input: validation, boundaries and review remain mandatory.