Kaenan, who sits in our Strategy Practice, published the book on November 10, 2026, after three decades where insurance strategy meets execution, building analytics organizations at USAA, Citibank, Sallie Mae, PwC and Ernst & Young, and building insurance businesses of his own. On launch day he described where the book came from:
“It grew out of the questions behind my work: What makes a good decision? What do experienced people know that never makes it into a manual? And as technology takes on more of the work, what should we be careful not to lose? I wrote it for insurance leaders, but I think the questions reach further than insurance. They are about people, responsibility and how we lead when having an answer is easier than knowing whether it’s the right one.”
We set out below the ideas from the book that we think matter most for leaders, in insurance and beyond.
AI changes the economics of judgment
Insurance is a judgment business. Every policy is a promise about an uncertain future, every underwriting decision is a judgment about risk, and every claim is a test of whether the promise meant what it said. Kaenan’s argument is that AI leaves this intact and changes the cost of everything around the judgment: for the first time, a carrier can strip large amounts of administrative work away from its experts and give them back time to think, to teach, to visit agents and to make better decisions.
That returned time is the Judgment Dividend, and it is paid only when leaders redesign the organization around it, which is why the book spends far more time on underwriters, claims professionals, agents, CEOs and Boards than on algorithms. The first of Kaenan’s Rules makes the same point: AI is not the competitive advantage; a disciplined culture that consistently makes better decisions is.
Ask which decisions AI has earned the right to make
Executives most often ask how much they should trust AI, and the book argues that this is the wrong question. The better one is which decisions AI has earned trust for, because trust in technology, like trust in people, is built over time and decision by decision.
Kaenan frames this as a Trust Ladder with five rungs:
- Observation. AI summarizes documents, finds relevant guidelines, highlights inconsistencies and suggests questions.
- Recommendation. AI proposes reserve ranges, flags underwriting concerns and suggests exceptions. People still make every decision.
- Assisted execution. Routine work, such as simple renewals and low-complexity endorsements, flows through automatically inside defined guardrails, with people supervising.
- Delegated authority. Some categories of decision run automatically because the organization has built enough evidence to trust the process, with monitoring, audit, appeal and escalation still in place.
- Organizational learning. Every decision, human or machine, makes the organization smarter. Claims improve underwriting, underwriting improves product and agents improve appetite.
The rule behind the ladder is that AI should never receive authority faster than the organization earns confidence in it, so each rung marks greater organizational confidence rather than greater technical sophistication.
Govern because people stay accountable
In the book’s framing, governance exists because people remain accountable for decisions even when technology helps make them, rather than because AI makes mistakes. Its test is simple and demanding: if a decision cannot be explained to the person it affected, the carrier has a governance problem it has not found yet, whatever the technology underneath. Kaenan writes from the boardroom here; his frameworks come from six consecutive years advising a regional mutual insurance group on innovation, and from AI steering committee seats at mutual carriers, where acceptable use policies and governance frameworks went from first draft to board practice.
Preserve what experienced people know
Much of what makes an underwriter or claims professional exceptional never made it into a manual. The book urges leaders to find out what made their experts good before those experts retire, and to preserve the thinking behind decisions as well as the decisions themselves. However, today’s experts learned judgment from work that AI is now removing, and if that apprenticeship disappears, leaders have to build a deliberate replacement for it; in the book’s words, build the school before you need the graduates. We see the same problem in heavy industry, and we set out our approach to it in Bridging the Skills Gap.
Build the company first, then choose the technology
One of the book’s rules will be familiar to anyone who has worked with us on technology strategy: don’t build an AI roadmap; build the company you want to lead, then choose technology that accelerates the journey. A related rule is to fund AI like a capability, not a project, because projects end and capabilities compound. We make the same case in mining, where a technology roadmap without an agreed destination rarely survives the first difficult budget, and our Mine of the Future guide sets out how we design that destination.
Never automate the human moment
Kaenan’s line is to automate everything that stands between the loss and the human moment, and never the human moment itself: the difficult claim conversation, the contested coverage call, the agency relationship that depends on someone’s judgment. Technology should take away administrative work and leave meaningful conversations with people, and while AI can improve confidence, it should never create the illusion of certainty.
Questions worth taking to the Board
The book closes each chapter with questions for the Board or executive committee. In the same spirit, these are the questions we would put to any leadership team adopting AI, in insurance or elsewhere:
- Which of our decisions has AI earned the right to influence, and on what evidence?
- Could we explain each automated decision to the person it affected?
- Where does our best expertise live today, and what happens to it when those people retire?
- Are we funding AI as a series of projects or as a capability we intend to compound?
- Which moments with customers, agents or partners must always stay human?
Inside the book
The Judgment Dividend runs to twenty-five chapters, written for leaders of regional and single-state carriers, though executives at national carriers will find plenty that applies to them. Each chapter carries recurring features built to be used rather than shelved: “Over a Bourbon” scenes, where executives say what they actually think after the meeting; Kaenan’s Rules, one idea per chapter worth keeping in view; Board Questions; and Leadership Reflections. The book is available in Kindle, paperback and hardcover editions.