The Challenge
Students can use AI but cannot evaluate it. Each time they accept output without evaluation, they accumulate cognitive debt (Kosmyna et al., 2025, MIT Media Lab). The debt compounds until they lack the capacity to catch AI errors.
The SAIL Framework
S
Social Intelligence — Communicating about AI to colleagues, clients, stakeholders
A
AI Literacy — Understanding AI capabilities, limitations, biases, and failure modes
I
Innovation/Inquiry — Questioning AI outputs; treating AI as thought partner, not oracle
L
Leadership — Taking responsibility for Human-AI collaboration outcomes
The Key Insight
Critical thinking isn't a fifth pillar—it's what makes the other four function.
The Greek kritikos means "able to discern." The APA Delphi Report (Facione, 1990) defines CT as "purposeful, self-regulatory judgment"—integrating analytical skills with dispositional qualities like open-mindedness and truth-seeking.
The Evidence: Three-Phase Scaffold
- 1.Foundation: Students own intellectual work; AI constrained to research assistance
- 2.Integration: Full AI use with required evaluation for hallucinations
- 3.Leadership: Design Human-AI collaboration frameworks as CSO
"The more I was depending on it, the less I was thinking about the concepts of underlying strategies."
— Student reflection, Strategic Management, Fall 2025 (Rosen)
The Promise
Skills can be automated. Judgment cannot.
Not just AI-literate graduates. Graduates with judgment.