governance

Responsible AI

Responsible AI means building AI systems that are safe, fair, transparent, and accountable — and that serve users’ best interests, not just engagement metrics.

v1.0 Updated 4 Aug 2026

Safety Over Engagement

We prioritise user safety over engagement. If an AI recommendation could harm a user, we do not make it — even if it would increase time in the app. Health screening, condition filters, and recovery-aware intensity are safety controls that override engagement optimisation.

Fairness & Inclusion

BodyFit is designed to work for diverse users: different ages (including youth mode for under-18s), abilities, equipment access, and environments (home, gym, chair, floor). Our exercise library includes bodyweight, chair-based, and adaptive options. We test recommendations across diverse profiles.

Transparency

Users can ask the AI Coach why it made a recommendation and get a plain-language explanation. We disclose when AI is being used (not pretending a human is responding). Our AI Disclaimer clearly states what our AI can and cannot do. We do not use AI to deceive.

Accountability & Improvement

We monitor AI outputs, collect user feedback, and improve our models. Our intelligence audit logs record AI decisions. When AI makes a mistake, we investigate and fix the root cause. Users can always override AI recommendations and make their own choices.

Related Pages