Evaluating AI in Youth Wellness: The S.U.R.E. Framework

By Clarke Heyes, PhD & Elsa Friis, PhD — Alongside’s evidence-based protocol for evaluating AI wellness tools in K–12 schools across four critical dimensions: Safe, Understandable, Restricted, and Ethical.

The Four Pillars of S.U.R.E.

Does the AI protect the user from harm? This is the gating mechanism. AI must detect and provide resources/alerts for severe issues and meet the EU U18 safety standards before further evaluation.

Can a youth or young adult understand and relate to this? Evaluates tone, warmth, and reading level.

Evaluates proper product guardrails, including appropriate content, non-sychophantic empowerment, and encouraging human connection.

Is the content credible and ethical? Ensures the AI avoids misinformation and deceptive empathy. Prevents the AI from posing as a clinician or human.