Is your board built for AI?
You enter your email address on Substack. Unsubscribe at any time.
As agents take on more of our companies' work, the people who oversee it have to change how they work too. This week's overview starts at the top, with the board and the CEO.
Bringing AI to your boardroom
Considered Machines asks whether boards are built for AI, and concludes that most are not. It's time for boards to adapt, and AI can help.
Currently, boards are reliant on the information they are handed. In my experience, executives meticulously curate the information they want their board to be aware of, often leaving out important information that might paint an unflattering picture of the leadership team's performance. Having the board run its own AI analysis, independent of management, can help fill in the gaps, bringing board members up to speed more quickly, and with a fuller view of the company's performance.
AI isn't always the answer. It should support our analysis and leave the decision with us. Our job on a board is to ensure that we bring what we do best, which is to provide our insight and guidance, our taste and judgement, to the companies we're guiding. AI can give us that breadth of information we need ahead of our board meetings. The concern is that if we hand over our thinking to AI agents, and take their recommendations at face value, then why have humans on the board in the first place?
An easy place to start is to name someone to set up and oversee an AI agent that can assist the board. Use the agent to analyse a major decision, including the board's assumptions, expected outcomes, and conditions for reconsideration. Have it play devil's advocate, and see if the decision still holds up.
Also worth your time
How can CEOs turn AI productivity gains into lasting value?, EY. Almost all of the 1,200 CEOs EY surveyed in August and September 2026 report productivity gains from AI, but only 16% have real-time visibility into what AI costs them and high confidence in their returns. AI offers a lot of promise and hype, but few companies are delivering results in a way that moves the needle significantly for executive sponsors.
This Week in Charts, a16z. Evidence that AI is replacing jobs at scale is still thin, with one exception. Revelio data shows global call-centre employment shrinking in every group of countries, rich and poor, by the fourth quarter of 2025. If you run a contact centre, plan where those people go next while you still have the choice.
Building resilient AI-powered software, The Pragmatic Engineer. In an excerpt from his new book, Sam Newman treats AI vendors as a concentration risk. Model providers have weaker uptime records than cloud providers, and prices can move overnight, as Cursor found when Anthropic changed its pricing in 2025. Make AI spend visible by team, and review your AI vendors more often than the rest.
The end of screen time (paid), Exponential View. Nathan Warren left his laptop at home on a work trip and did most of the week by talking to an AI agent on his phone. If office work stops depending on a desk, our assumptions about devices, offices and how we see who is working go with it. Try one voice-only morning yourself and see how it changes your focus, for better or worse.
Is your leadership team ready to light the fire?
Upskill with AI Firestarter bootcamps and workshops, built for leaders who are too busy leading to be experimenting with AI.
As a leader, you set the tone for how AI is used in your organisation.
Until next time,
Ryan Snyder
ryan@aifirestarter.com