AI at Measure exists to strengthen equity - not replace humanity.
How we govern AI on this website and across our work: principles, disclosure, the policies we follow, and how to report harm.
Our principles
Lines in the sand.
Community accountability over technological efficiency
If a tool cannot equitably represent community experiences, we challenge its validity rather than accept its conclusions.
AI Interrogation, always
We scrutinize data sources, algorithms, and decision models using our Theory of Interrogative Reasoning to surface embedded bias and structural inequity.
Communities in the loop
AI may assist analysis. It cannot replace the interpretation of lived experience. Communities review every public-facing AI output.
Data dignity
PII, PHI, participant data, and confidential organizational data never enter unapproved AI systems. Period.
Disclosure at first touch
We tell you when you're interacting with an AI-powered tool - at or before first interaction, in plain, accessible language.
Environmental & systemic responsibility
We name AI's energy and water footprint, and how data center siting disproportionately burdens frontline BIPOC communities.
Where AI shows up on this site
Full transparency about our AI surfaces.
Per Section 13A of our policy, we disclose AI use before first interaction. Below is every place AI runs on wemeasure.org, what it does, and what data it sees.
Engagement Finder
- Where
- About → Join us in driving change
- Purpose
- Recommends Measure pathways (donate, Free Data Support, services, learn) based on the goals you select.
- Model
- Google Gemini via Lovable AI Gateway
- Data sent
- Only the form fields you submit - no account, no tracking, no PII required.
Permitted uses
All require human review before publication or decision.
- Drafting reports, training materials, and communications (with human review)
- Marketing copy, social posts, and graphics
- Research summaries and synthesis support
- Early drafts of grant proposals
- Meeting note drafts
- Data exploration (not final decision-making)
- Navigation bots and platform tools
- Generating community-co-designed training data
Prohibited uses
Items marked TRAIGA are also prohibited under Texas law.
- Determining participant eligibility or automating program outcomes
- Entering PII, PHI, or sensitive personal data into unapproved systems
- Publishing AI-generated content without human review
- Manipulation, misinformation, or bypassing ethical oversight
- Surveillance of staff productivity
- AI designed to incite self-harm, harm to others, or criminal activity (TRAIGA)
- AI used to unlawfully discriminate based on a protected class (TRAIGA)
- AI social scoring resulting in detrimental treatment (TRAIGA)
- Biometric identification from public sources without consent (TRAIGA)
- Agentic AI without documented ELT approval and human review
Compliance & alignment
Governed by the highest standards in the field.
Texas Responsible AI Governance Act - Measure operates as a deployer.
Healthcare AI provisions, effective Sept 1, 2025.
Federal AI executive order (Jan 23, 2025).
Responsible governance of emerging technologies.
Our compliance roadmap and TRAIGA safe-harbor anchor.
Operations produces a NIST alignment summary at every quarterly review as ongoing proof of compliance and TRAIGA safe-harbor positioning.
Incident response
When AI causes harm, we move fast.
Appendix A of our policy is the AI Incident Response Plan (AIRP). It governs detection, triage, containment, correction, and learning.
Submit an incidentNotify ELT, contain incident.
Preliminary assessment and documentation.
Corrective action plan and communication strategy.
Final Incident Report to ELT, update AI Tool Log.
When AI conflicts with community accountability or staff well-being, community and well-being take priority.
- Conclusion, Measure AI Acceptable Use Policy, V1 · March 2026