Leaders Must Understand What They Govern

A governing body receives a technical briefing. The experts speak with confidence, the diagrams are complex and the consequences appear urgent. Some leaders defer because they do not want to expose what they do not know. Others reject the advice because they distrust language they cannot test. A third group repeats the expert conclusion as though receiving it had completed the work of governance.

None of these responses is sufficient. Leaders cannot master every discipline represented in a modern institution, and pretending otherwise creates its own danger. But they remain responsible for purpose, authority, resources, rights, trade-offs and consequences. Expertise can inform that responsibility; it cannot absorb it.

Leaders must understand what they govern. The required understanding is not professional mastery of every method. It is decision-grade understanding: enough knowledge to frame the real question, test the evidence, identify interests and uncertainty, hear people outside the expert room, compare credible alternatives, set conditions, and explain why the authorised decision was made.

On 23 September 2026, leaders of major artificial-intelligence companies and independent experts briefed the United Nations Security Council on AI and international security. Reuters reported the meeting at 10:05 UTC. Anthropic chief executive Dario Amodei warned that poorly managed AI could threaten humanity; OpenAI chief executive Sam Altman called for international cooperation; and Yoshua Bengio, co-chair of the United Nations Independent International Scientific Panel…