alexe-ev Govern Responsible AI
Purpose
Help product teams establish governance practices that ensure AI features are developed and deployed responsibly — with fairness, transparency, accountability, and safety built in.
Skill type
Conceptual skill
Use this skill when
- An AI feature is being developed and governance standards need to be defined
- Existing AI features lack accountability or monitoring mechanisms
- A team wants to establish responsible AI principles before building
- Regulatory or stakeholder pressure requires a responsible AI framework
Do not use this skill when
- The goal is AI quality monitoring only (use evaluate-ai-quality-monitoring)
- The goal is ethical review of a specific decision (use apply-ethical-decision-framework)
Required inputs
- AI feature or product being governed
- Organizational context (team size, AI maturity, industry)
Optional inputs
- Existing AI ethics principles
- Regulatory context (EU AI Act, sector-specific rules)
- Known AI risks in the product area
- Stakeholder expectations
Upstream context
Works best when:
- AI feature is defined
- Ethical and privacy assessments have been done
Downstream handoff
Output can feed:
- evaluate-ai-quality-monitoring (governance standards feed quality criteria)
- assess-privacy-security-regulatory (AI governance intersects with privacy)
Instructions
- Define the AI governance principles for this context (fairness, transparency, accountability, safety, privacy).
- Identify governance requirements per principle.
- Design accountability structures: who owns AI governance, how decisions are made.
- Design transparency mechanisms: how users are informed about AI use.
- Define fairness criteria and assessment methods.
- Design audit and review processes.
- Define escalation paths for AI governance failures.
Output
Provide:
- AI governance principles (tailored to context)
- Requirements per principle
- Accountability structure
- Transparency mechanisms
- Fairness criteria and assessment approach
- Audit and review process
- Escalation paths
- Governance maturity level: minimal / basic / mature
Risks / caveats
- Governance principles without enforcement are decoration
- Fairness requires defining "fair for whom" — it's context-specific
- AI governance must involve legal, ethics, and technical stakeholders, not just product