9. Accountability & Governance in Practice

Living worlds evolve, player expectations shift, and new risks emerge over time.

9.1 Roles & Accountabilities

Each Of Moss & Moonlight Commitment has an accountable owner responsible for delivery, oversight, and intervention where required.

These roles are capability-based - one person may hold multiple roles, and some capabilities may be supported by external advisors. Each capability has a named owner, even when responsibilities are shared.

Roles responsible for emotional safety and intervention require appropriate resourcing, support, and rotation to avoid secondary harm to those performing them. As Of Moss & Moonlight grows, these roles may expand, split, or consolidate - but the need for coverage and accountability does not change.  

From the paper:
Framework for
Responsible AI in Games

Full paper release
15th May 2026
Responsibility is not abstract. It is assigned, owned, and reviewable.

a) Governance Capabilities

  • Governance Lead - Ethics & Player Trust: Owns the Five + Five standard and key safety boundaries.
  • Governance Lead - Studio Policy & Risk: Owns release gates, policy compliance, change governance.
  • External Reviewer - Ethics & Player Safety: Independent review for high-risk systems or major AI shifts.

b) Design Capabilities

  • Design Lead - Emotional Safety: Emotional safety patterns, opt-outs, de-escalation.
  • Design Lead - Consent & Intensity UX: Consent controls, intensity settings, boundary revisiting.
  • Design Lead - Player Transparency & Trust UX: Player-facing disclosure, clarity, and comprehension.
  • Design Lead - Narrative Integrity: Tone, intent, authored constraints, and narrative alignment.
  • Canon Owner - Lore & Continuity: Lore truth sources, continuity, canon coherence.

c) Engineering / Runtime Capabilities

  • Runtime Lead - Safety & Intervention: Guardrails, interventions, drift monitoring, rollback/reset.
  • Engineering Lead - Data Protection & Privacy: Privacy-by-design, anonymized collection, memory clarity.
  • Engineering Lead - Helix Runtime: Runtime implementation, reliability, enforcement mechanisms.
  • Engineering Lead - Tooling & Observability: Tracing, dashboards, diagnostics, and repair tools.
  • Engineering Lead - Sustainability Optimization: Optimization, routing efficiency, footprint tracking.

d) Operations Capabilities

  • Ops Lead - Incident Response & Community Trust: Owns reporting workflows, triage, escalation, community feedback loops, and trustworthy follow-through.
Accountability is not a static structure - it is an ongoing commitment
to ownership, intervention, and stewardship in practice.

9.2 Join the Conversation

This work is shared as part of our broader Responsible AI practice.

If you have a perspective, challenge, or insight, we’d welcome it as part of the Five + Five conversation.

Join the Conversation →