AIP › Research › Evidence

Balanced Evidence. Harm + Potential.

Both are true. Harm without guardrails. Potential with guardrails. AIP enables both.

Documented Harm Cases.

60%
U.S. adults worry AI is under-regulated (Pew Research, April 2025, 5,410 adults)
78%
EU citizens want public authority oversight (Eurobarometer 554, ~27,000 citizens, 27 countries)

Success Cases with Guardrails.

18
Documented K–12 deployments reviewed (see case cards for sources) (Khanmigo, Amira, Quill.org, MagicSchool, TalkPal, Photomath, Diffit, ClassDojo)
18
Documented K–12 deployments with published guardrails. Improvement figures are the vendors’ own and are not independently verified — each case card states what was measured, by whom, and over how long.
64%
Teens use AI chatbots (Pew Research, Dec 2025). 21% workers use AI (growing workforce demand)

Key Insight

The evidence is clear: AI without proper guardrails causes real harm. But AI with appropriate policy frameworks enables remarkable success. The AIP framework provides the guardrails that allow innovation to flourish safely.