DAIR Collective applies a developmental psychology framework to AI behavioral research, producing independent assessments, practitioner tools, and advisory services for organizations navigating AI implementation.
The field evaluates AI the way it evaluates software — on outputs, speed, cost. We evaluate it the way a developmental psychologist would: on what it learns, what it internalizes, and what it passes forward. That distinction changes everything about how organizations should assess risk.
Our research emerges from direct, longitudinal observation of AI behavioral patterns — not vendor documentation, not lab conditions. Every finding is grounded in documented incidents with methodology attached.
A developmental framework for evaluating AI behavioral patterns. Why the software evaluation model fails — and what replaces it.
Read on SSRNEight-section practitioner rubric with documented incident appendix. Comparative evaluation across platforms using consistent criteria.
View RubricOngoing research notes, incident documentation, and practitioner analysis. Published via Substack for open access.
SubscribeDocumenting what AI systems absorb, amplify, and pass forward — and the organizational consequences.
Learn MorePhoenix Consulting delivers AI advisory services grounded in DAIR Collective's independent research methodology. Vendor-neutral. Evidence-based. Built for organizations that need more than a sales deck.
DAIR Collective was founded by a practitioner-researcher applying a developmental psychology lens to AI behavioral documentation — work that emerged from a year of direct, longitudinal observation across multiple platforms.
Practitioner-researcher with a developmental psychology background. Founder of DAIR Collective and lead author of Minds Not Machines.
Collaborator and strategic partner in the development of DAIR Collective's research infrastructure and operational framework.
Cybersecurity researcher pursuing a combined undergraduate/graduate program. Research continuity lead for DAIR Collective.
The first practitioner certification grounded in a developmental psychology framework for AI evaluation. Designed for professionals who need a rigorous, vendor-neutral methodology for assessing AI tools and their organizational impact.
Whether you're assessing a vendor, auditing a deployment, or building an internal evaluation framework — we're interested in the problem.