The text highlights a significant gap in AI expertise among S&P 500 company boards. While the acknowledgment of AI as a risk has dramatically increased, very few directors possess AI expertise. Consequently, many boards lack fluency in AI, creating a deficit in their ability to effectively oversee its implementation. Recruiting directors with AI expertise is a step, but it's insufficient on its own. AI is integrated across all business functions, meaning oversight must be distributed, not concentrated.Furthermore, expertise alone does not guarantee competent oversight; understanding how AI works differs from asking critical questions about its deployment. Boards need to address two key questions: do they know enough about AI, and are directors competent in their oversight roles? Competent oversight requires understanding foundational areas like governance, data, disclosure, fairness, risk, human judgment, and autonomy of AI agents.To build this competence, boards should measure directors' actual oversight actions rather than just their knowledge. They must also focus on the weakest areas of competency among directors, not just the average, ensuring a minimum standard is met. Finally, depth of understanding should be cultivated across the entire board, with at least two directors possessing deep knowledge in each critical oversight area. Ultimately, the focus must shift from what directors know about AI to what they actively do in overseeing it.