Four Questions Every Leader Should Ask Before Building an AI Strategy

Most organizations feel pressure to have an AI strategy. Far fewer know where to begin. A new piece from the Stanford Social Innovation Review offers a framework that applies well beyond nonprofits: four foundational questions that leaders should answer before picking tools, hiring talent, or launching pilots.

The Four Questions

1. AI for What?

Before selecting tools, clarify the problem. The authors suggest mapping AI opportunities along a spectrum — from incremental productivity gains (automating admin, drafting content) to transformational ambitions (expanding reach, improving outcomes at scale). Tools come later. Purpose comes first.

2. What Will It Take?

AI readiness depends on three dimensions: your organization’s relationship with technology (tech-lite vs. tech-forward), its learning culture (does it use data to adapt or just to report?), and its operating domain (some fields carry higher data quality or regulatory complexity). Honest assessment here prevents expensive misfires.

Key takeaways:

• Start with AI literacy, not hiring. The authors recommend building broad internal AI fluency before investing in technical talent — a staged “dosing strategy” that avoids over-indexing on one role too early.
• AI adoption is not linear. It requires iteration and a willingness to change direction. Organizations with strong learning cultures are better positioned to explore AI responsibly.
• Pace is a strategic choice. Moving at the “speed of trust” — shaped by inclusion commitments, safeguards, and compliance — is not a sign of fear. It’s a sign of strategic discipline.

3. How Do We Lead Through This?

Acknowledge organizational tension openly. Establish responsible AI principles early — covering fairness, transparency, data privacy, and human oversight — to give teams a shared ethical framework for decisions.

4. What Pace Is Right for Us?

The private sector frames AI as a speed race. Mission-driven organizations operate under a different mandate, where a flawed deployment can cause real harm. Articulating why pace is intentional, not fearful, matters.

🔗 Read the full article on Stanford Social Innovation Review