M. Sohaib Khan’s Agile in the Age of AI is a practical guide for leaders trying to coordinate complex work across many teams. Its central argument begins with a familiar problem: as the number of teams grows, dependencies multiply, while each person’s capacity to track them remains limited. Khan proposes using AI to spot risks, synthesize information, and bring emerging problems to leaders’ attention sooner. He frames Agile as a way to learn and adapt, then places AI within that process as a means of detecting signals at a wider scale. This gives the book a clear purpose: helping organizations make delivery more predictable while preserving their ability to respond to change.

The book’s main contribution is Khan’s Adaptive Governance Delivery Framework, which connects predictive governance, adaptive delivery, and continuous alignment. Charters, outcome maps, decision rights, investment guardrails, and review points give leaders a defined way to act on information surfaced by AI. Khan is especially effective when he specifies who owns a decision. “A role is defined by what it can decide, not by what it does,” he writes, a principle that gives his discussion of governance practical force. His examples also keep attention on business outcomes: a team’s success depends on whether its work improves a customer or organizational result, not simply on how many features it completes.

Much of the book is designed for use at work. Separate playbooks address release train engineers, product owners, Scrum Masters, executives, technical leaders, and change leaders. They’re joined by prompts for planning, dependency management, retrospectives, risk assessment, and difficult decisions, as well as templates and a detailed 90-day adoption plan. The chapters on fear, resistance, and psychological safety give this material necessary human context. Khan’s recurring rule, “AI advises, humans decide,” becomes more concrete in his descriptions of human review, documented accountability, and the need to question an AI recommendation that lacks context. The anonymized case studies illustrate how these practices might fit together across organizations of different sizes.

I think the book is strongest as a working reference that readers can adapt to a particular team or program. Its abundance of prompts and checklists makes it easy to find a starting point, though repetition can slow a continuous reading. Khan explains that the case studies are composites and that their figures are approximate; readers should apply the same care to the book’s ambitious productivity and return estimates when building their own business cases. The different timelines for introducing AI also call for judgment: the rapid 90-day journey and the more cautious maturity path suit different starting conditions. Even with those qualifications, Agile in the Age of AI offers a useful, detailed account of how leaders can pair better information with clear responsibility, purposeful meetings, and sustained attention to the people doing the work.