10. Applied Practice & Open Questions
What it looks like in practice - and where it’s still unresolved.
10.1 Implementation Realities
This section shows how Vinebright applies the Five + Five framework in practice - through Helix as a living world runtime, and through specific design, governance, and operational commitments for Of Moss & Moonlight. This is one application example, shaped by our values and constraints.
Building the Of Moss & Moonlight prototype has been equal parts exhilarating and humbling. Some things worked sooner than expected; others forced us to slow down, rethink assumptions, and redesign with more care. That is not a failure of the framework - it is the reality of building living worlds with real people inside them.
Responsible AI in Games
15th May 2026
This framework does not claim to resolve all ethical, psychological, or regulatory questions associated with AI in games. It does not prescribe a single “correct” implementation, nor replace legal, platform, or player-specific safeguards.
Implementation surfaces trade-offs that no framework can fully resolve.
- Designing consent mechanisms that remain meaningful without becoming disruptive.
- Balancing emotional richness with safety and legibility.
- Introducing guardrails without flattening character, tone, or narrative intent.
- Maintaining observability and accountability without introducing surveillance.
Some questions remain unresolved.
- How trade-offs across safety, coherence, consent, and sustainability are balanced with model memory.
- How player expectations will shift as living worlds become more common.
- How responsibility changes as systems move toward sustained, high-intensity interaction.
- What additional roles, review practices, or safeguards may become necessary over time, and how and when is consent re-requested.
These questions do not have universal answers. Appropriate responses depend on genre, audience, culture, art direction, and community norms. What is proportionate in one context may be unnecessary, or insufficient, in another. Studios may draw on this framework in whole, in part, or not at all.
Living worlds require ongoing attention. Responsibility does not conclude at launch, nor does it sit solely in documentation or diagrams. It is maintained through observation, intervention, review, and revision as systems evolve.
This work is continuous, contextual, and shared - and it will continue to change
as these systems, and the expectations around them, mature.
10.2 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.
