What is the Authentic Intelligence Model?

Start with the decision and the people affected

AIM begins before selecting a product or accepting a vendor’s promise. Define the decision, identify whose opportunities or interests it affects, and describe the role AI would play. Administrative support, decision support, and decision automation create different demands on human judgment and accountability.

The model grew from doctoral research on AI-mediated hiring in financial services. The book extends its questions to other consequential settings. The organizing question is practical: can the organization own this AI-shaped decision, including its consequences when the system is wrong?

Gate 0: should AI be used at all?

Gate 0 is AIM’s permission-to-use decision gate. Apply it before procurement, deployment, or expansion to ask whether a defined AI use can meet the organization’s ethical obligations.

Its five threshold checks examine role definition; baseline duties and data stewardship; inequality and lifecycle impacts; accountability and moral agency; and organizational and vendor readiness. In the hiring research, protections include nondiscrimination, accessibility, privacy, meaningful notice, intelligible explanation, and recourse.

These protections are threshold duties. An efficiency gain cannot, by itself, justify removing a person’s ability to understand or challenge a consequential decision. Missing safeguards may require declining the use, narrowing its role, or making adoption conditional on changes that can be verified.

The Triangle: Rights, Reasons, and Results

The Triangle examines three stakeholder relationships. In the dissertation these are the applicant, the hiring manager, and the institution-community. In wider applications, leaders identify the affected person, the decision professional, and the institution or community.

Rights

Are baseline protections preserved in the actual workflow? Can the affected person obtain notice, accessible treatment, and meaningful recourse?

Reasons

Can the organization give intelligible, relevant reasons for the decision? Can a person challenge those reasons and obtain a response from someone with authority?

Results

What happens to people over time? Are outcomes monitored, harms investigated, and conditions for changing or stopping use enforced?

All three questions apply across the stakeholder relationships. A performance metric alone cannot establish that the whole decision process is legitimate.

The Orbit: account for wider effects

The Orbit widens the review beyond the immediate decision. It considers workforce effects, environmental stewardship, security, resilience, community trust, and dependencies within the wider vendor and platform ecosystem.

Vendor governance runs across Gate 0, the Triangle, and the Orbit. A vendor is not a fourth corner of the Triangle. Its control over data, model updates, explanations, monitoring, and correction can determine whether the institution can meet its obligations at all.

How can a leadership team use AIM?

1. Name the decision. Record the proposed AI role, the people affected, and the accountable owner.
2. Review permission. Use Gate 0 to identify missing protections and the evidence needed before use can proceed.
3. Examine relationships and wider effects. Work through the Triangle and Orbit, including what the vendor enables or prevents.
4. Record and revisit the judgment. Make conditions, owners, appeal pathways, monitoring thresholds, and stop authority explicit.

The free companion workbook contains AIM on One Page, a Gate 0 worksheet, Triangle and Orbit scorecards, and decision records. Use the materials at procurement, pilot, deployment, renewal or vendor change, and after an incident, complaint, or drift signal. No book purchase is needed to request the workbook.