Privacy by default
Never enter identifying information about children, young people, participants, staff or safeguarding cases into any AI tool. Replace names with roles, aggregate wherever possible, and treat every prompt as if it may be logged.
Safeguards & guidance
Non-negotiables before you use AI at the YMCA.
The YMCA has built trust across communities for over 175 years. Generative AI is a powerful colleague — but only when it operates inside the same ethics that make our movement worth trusting.
01 — Principles
Six principles for using AI responsibly.
Apply these to every prompt, every output, every share.
Youth voice protection
AI must never speak for young people. Use it to prepare listening sessions, summarise consented feedback, or draft accessible resources — not to synthesise opinions that were never expressed.
Human accountability
AI drafts, humans decide. Every output benefits from a named reviewer who checks accuracy, tone, and fit with local context. Safeguarding, funding and policy decisions always require human sign-off.
Cultural and local context
The YMCA operates in 120+ countries. Generic outputs can flatten difference. Anchor every prompt in your community, language, and priorities — and edit outputs that don't fit.
Verify, don't trust
AI is confident even when it's wrong. For statistics, dates, references or policy specifics, verify against a trusted source before publishing or sharing externally.
Transparency with those affected
When AI has meaningfully shaped a public-facing communication, be honest about it. Trust is our currency — and it compounds through openness.
02 — Red lines
Never share these with AI.
Copy this list to your team. If in doubt, don't paste it.
- ×Identifiable information about a child or young person
- ×Health, safeguarding, disciplinary or immigration information about any individual
- ×Confidential board, HR or partnership material without permission
- ×Content that could substitute for professional advice (legal, medical, therapeutic)
- ×Anything you would not be comfortable seeing on a screenshot
03 — If it goes wrong
When something goes wrong
If an AI output causes harm, feels unsafe, or breaches privacy:
- 1Stop and do not publish or share the output further.
- 2Tell your line manager or designated safeguarding lead.
- 3Follow your association's existing incident and safeguarding process — AI does not change that.
- 4Share the learning so others can avoid the same mistake.
04 — Next
Ready to try a prompt?
Every prompt in the library has these safeguards embedded in its checklist.
Open the prompt library