Keeping up with artificial intelligence has become almost a full-time job. Every week brings a new model, a new term, an acronym that did not exist yesterday and feels indispensable today. It is easy to feel a step behind, even for people who work in tech.

Summer, though, offers one of the rare chances to slow down and make sense of it all: to read about AI while stepping away, for a few hours, from AI itself. A good book does what no feed can, slowing the present down just enough to make it legible.

So here are ten books, plus a must-read bonus, to pack in your bag. The next time someone by the pool claims that "AI will make every decision for us", you will not be caught off guard: you will have the vocabulary and the arguments to answer.

This year's releases

1. Empire of AI, by Karen Hao (Penguin, 2025)

Inside the AI factory.

The most thoroughly reported chronicle of the industry reshaping every market, told by the first journalist to get inside OpenAI. Hao, formerly of MIT Technology Review and The Atlantic, reconstructs the ambitions, capital and power dynamics of the AI race. For a decision-maker, understanding the system that builds these tools matters as much as understanding the tools.

2. Co-Intelligence, by Ethan Mollick (Portfolio, 2024)

Italian edition: "L'intelligenza condivisa" (Luiss University Press, 2025).

AI as a colleague, not an oracle.

The most pragmatic guide to working with generative AI: where to delegate, where to collaborate, where to be sceptical. Mollick, a Wharton professor, turns months of experimentation into simple, testable principles. No prophecies, just daily practices for teams and managers, the ideal bridge from excitement to real use.

3. Forma mentis, by Nello Cristianini (Il Mulino, 2025) — Italian only

What happens inside the models.

A journey into interpretability, the field trying to make neural networks legible and verifiable. Cristianini, a machine-learning pioneer in Europe, describes precisely what research has understood and what remains opaque. Trust, governance and accountability over automated decisions start right here. One of the most compact and original Italian essays of the year.

4. AI Snake Oil, by Arvind Narayanan and Sayash Kapoor (Princeton University Press, 2024)

Telling real AI from snake oil.

The essential anti-hype tool for anyone assessing vendors and investments: what AI can really do versus what it is merely promised to do. Narayanan, a Princeton professor, and Kapoor dismantle market exaggerations with rigour and teach you to ask for the right evidence before signing. The perfect counterweight to books obsessed with superintelligence.

5. The Coming Wave, by Mustafa Suleyman and Michael Bhaskar (Crown, 2023)

Italian edition: "L'onda che verrà" (Garzanti, 2024).

Govern the wave, don't just ride it.

A strategic reflection on the twin wave of AI and biotech, and on how to contain its risks without giving up the opportunities. Suleyman, DeepMind co-founder and now CEO of Microsoft AI, combines technical depth with a systems view. A book written for those who lead organisations and must decide now: ambitious, contested, necessary.

The method: causal AI and decision intelligence

6. Causal Artificial Intelligence, by Judith S. Hurwitz and John K. Thompson (Wiley, 2023)

AI that understands why, not just what.

The book that brings causal AI into the enterprise: why correlation-only AI is not enough, and how causal models improve decisions, with concrete use cases. Hurwitz (chief evangelist at a causal AI platform) and Thompson (head of global AI operations at EY) write as practitioners, not academics. The subtitle says it all: the next step in effective business AI.

7. Link, by Lorien Pratt (Emerald Publishing, 2019)

Start from the decision, not the data.

The book that founded decision intelligence: how to connect data, actions and outcomes by putting the decision at the centre. Pratt effectively defined the discipline Vedrai's work rests on, offering a method to turn models and predictions into better, measurable choices. For hands-on use, her "Decision Intelligence Handbook" (2023) is the ideal companion.

8. Power and Prediction, by Ajay Agrawal, Joshua Gans and Avi Goldfarb (Harvard Business Review Press, 2022)

Italian edition: "Potere e previsione" (Franco Angeli, 2024).

The economics of decisions in the age of AI.

The most cited economic thesis: AI lowers the cost of prediction, and that changes the value of every decision. The three University of Toronto economists explain why the real leap is not individual tools but the redesign of decision systems. A perfect bridge between business economics and decision intelligence, best read alongside their classic "Prediction Machines".

The long view

9. Nexus, by Yuval Noah Harari (Fern Press, 2024)

Whoever controls information, rules.

A history of information networks, from print to algorithms, showing how data flows shape societies. Harari, the world's most-read historian on big themes, provides the wide frame within which to place AI. Not a technical book, but a map of the terrain on which our decisions take shape, useful for lifting your gaze beyond the single use case.

10. The Ethics of Artificial Intelligence, by Luciano Floridi (Oxford University Press, 2023)

Italian edition: "Etica dell'intelligenza artificiale" (Raffaello Cortina, 2022).

The rules of the game.

The reference framework on governance, responsibility and the correct use of AI, by the most internationally influential Italian philosopher of information. Floridi translates abstract principles into criteria organisations can apply, just as the rules (from the AI Act onward) enter daily practice. Because deciding well with AI also means deciding legitimately.

The must-read bonus

The Book of Why, by Judea Pearl and Dana Mackenzie (Basic Books, 2018)

The roots of causal AI.

A few years old, but for us at Vedrai it remains a must-read: the work that popularised the science of cause and effect. Judea Pearl, Turing Award winner and father of causal inference, explains why machines must learn to answer the question "why?". It is the theoretical root beneath today's causal AI and decision intelligence, and the natural prologue to Hurwitz and Thompson's book. If a single title had to stay in your bag next summer too, this is the one.

In closing

Eleven books will not turn anyone into an expert in two weeks. They do something more useful: give you a vocabulary and a method to tell what matters from what is merely noise. That is exactly the work we do every day at Vedrai with decision intelligence: helping companies decide better, not delegate their decisions away. The next time someone by the pool claims that "AI will decide everything for us", you will have the tools to answer. Happy reading, and happy summer.

Frequently asked questions

What are the best AI books to read in 2026?

Top picks include Empire of AI by Karen Hao, Co-Intelligence by Ethan Mollick, Forma mentis by Nello Cristianini, AI Snake Oil by Narayanan and Kapoor and The Coming Wave by Mustafa Suleyman, alongside the key references on causal AI and decision intelligence.

What is the go-to book on causal AI?

For business, it is Causal Artificial Intelligence by Judith Hurwitz and John Thompson (2023). The foundational classic, still essential, is The Book of Why by Judea Pearl, Turing Award winner and father of causal inference.

What is the essential book on decision intelligence?

The book that founded the discipline is Link by Lorien Pratt, on connecting data, actions and outcomes by putting the decision first. For practical use, her Decision Intelligence Handbook is a great companion.

Are there AI books available in Italian?

Yes: Forma mentis by Nello Cristianini, Co-Intelligence (as "L'intelligenza condivisa"), The Coming Wave ("L'onda che verrà"), Nexus by Yuval Noah Harari, Power and Prediction ("Potere e previsione") and Floridi's Ethics of Artificial Intelligence.