Feeling overwhelmed by the AI doom loop? Read this reading list designed for decision‑makers
If your inbox swings between existential alarms and investor euphoria, a curated set of books is the fastest way out of the noise. Reporters, academics and some practitioners point to the same cluster of titles, including investigative reporting, polemics, environmental and historical analysis, and even fiction. These help leaders separate immediate, empirical risks from longer‑term theoretical concerns.
Shortlist for the C‑suite, read these three first
- Empire of AI, Karen Hao. Why it matters: a primer on industry dynamics, product choices and reputational fault lines for CEOs and boards.
- Automating Inequality, Virginia Eubanks (2018). Why it matters: concrete examples of how automated systems produce legal, social and compliance risk, must‑read for legal, procurement and public‑sector partners.
- Hyperscale, Paris Marx. Why it matters: maps the environmental and community impacts of data‑center expansion; essential for sustainability and real‑estate planning.
Hold two distinctions
Think in two buckets:
- Long‑term, alignment and philosophical risk. These works, often speculative and model‑oriented, ask whether machine intelligence could someday cascade into catastrophic outcomes. They include Nick Bostrom’s Superintelligence (2014) and writings by Eliezer Yudkowsky and Nate Soares.
- Near‑term, systemic harms and power dynamics. These books document how AI is already reshaping surveillance, schooling, policing, energy, and labor, places where executives can and should act now. Examples include Virginia Eubanks, Kashmir Hill, Paris Marx and Anita Say Chan.
Both matter. For business leaders, act now on verifiable near‑term harms and study long‑term arguments so governance and strategy are stress‑tested against extreme scenarios.
Essential reads, grouped, annotated, and triage‑ready
Investigative reporting on industry and deployments
- Empire of AI, Karen Hao. Intensive reporting on the industry’s rise, including coverage of OpenAI, the formation of Anthropic, and public ruptures among tech leaders (early reviews in 2025 called it overly negative, and the reading list notes she has since been “entirely vindicated”). Why it matters: shows how fundraising, product design and internal culture produce risk vectors that leak to customers and regulators.
- Project Maven, Katrina Manson. Focuses on U.S. military collaborations and the use of industry tools for surveillance and autonomy. Why it matters: essential context for companies considering government or dual‑use contracts, and for compliance teams vetting export and procurement risk.
- Your Face Belongs to Us, Kashmir Hill. Reporting on Clearview AI and how biometric databases spread through law enforcement and private sectors. Why it matters: a cautionary tale about privacy, vendor risk and downstream reputational exposure.
Controversial or polemical perspectives (read to understand their logic)
- The AI Con, Emily M. Bender & Alex Hanna. A pointed critique of exaggerated claims about imminent human replacement and of commercial rhetoric that can justify broad data extraction and surveillance. Why it matters: helps procurement and product teams translate vendor marketing into measurable claims and audit requirements.
- If Anyone Builds It, Everyone Dies, Eliezer Yudkowsky & Nate Soares. A representative long‑term alignment manifesto, dense with thought experiments and worst‑case logic (the reading list labels it a “doomer classic”, and one review noted it “reads like a Scientology manual”). Why it matters: use it to design tabletop exercises that stress governance under extreme failure modes.
- Genesis, Henry Kissinger, Eric Schmidt, Craig Mundie, Eleanor Runde (2024). Argues that opaque AI decision‑making threatens the “age of reason” and that companies owning AI could amass sweeping power, or that AI might become a “philosopher king.” Why it matters: clarifies how policymakers and elites are framing the stakes at the highest level.
Context, history, environment and fiction that sharpen judgment
- Hyperscale, Paris Marx. Examines the physical footprint of AI growth: data centers, power and water use, and local social impacts. Why it matters: informs site selection, supplier due diligence and sustainability KPIs.
- Predatory Data, Anita Say Chan and Automating Inequality, Virginia Eubanks (2018). Chan links modern techno‑surveillance to historical patterns of exclusion, and Eubanks documents how automated systems entrench inequality. Why it matters: both supply procurement‑level red flags, who benefits, who is harmed, and what legal exposure follows.
- Prophecy, Carissa Véliz. “The main promise of prediction is not knowledge of the future but domination over others.” Why it matters: reframes predictive systems as political instruments as much as optimisation tools, important for public affairs and ethics reviews.
- Fiction and thought experiments: I, Robot, Isaac Asimov (1950); Labyrinths, Jorge Luis Borges; The Lost Books of the Odyssey, Zachary Mason (2007). Why it matters: Asimov’s Three Laws provide a shared shorthand for safety conversations. Borges and Mason surface questions about language, authorship and how LLMs reconfigure meaning, useful when designing human‑in‑the‑loop controls and content governance.
Books about people and ideologies shaping tech culture
- The Technological Republic, Alex Karp & Nicholas Zamiska. A look into certain CEO mindsets and arguments for concentrated technical authority (the reading list includes sharp stylistic critiques). Why it matters: gives insight into how top leaders justify centralised decision rights and the governance tradeoffs that follow.
- More Everything Forever, Adam Becker and The Rise and Fall of the Artificial State, Jill Lepore. Why they matter: trace the ideological lineages, accelerationism, rationalist communities and political economies that shape product priorities and risk appetites across the industry.
Direct quotes worth handing to your board
“could kill us all by the end of the decade.”, a recently resigned AI company employee (reported 2026/sep/15).
“the main promise of prediction is not knowledge of the future but domination over others.”, Carissa Véliz.
“A robot may not injure a human being or, through inaction, allow a human being to come to harm.”, Isaac Asimov (Three Laws of Robotics).
Key questions a curious executive will ask (and short answers)
-
Which books help me spot immediate, actionable risks to my business?
Start with investigative and policy‑focused reporting: Empire of AI for industry behaviour and reputational patterns; Automating Inequality and Predatory Data for discrimination and procurement risk; Hyperscale for environmental and infrastructure exposure.
-
Which works explain the long‑term “alignment” or existential concerns and should I treat them as operational priorities?
Read Superintelligence, and writings by alignment proponents like Yudkowsky & Soares to understand worst‑case scenarios. Treat their scenarios as stress‑tests for governance rather than immediate operational checklists, and incorporate relevant mitigations into high‑severity incident playbooks.
-
How do I separate hype from real capability when vendors claim breakthroughs?
Translate marketing into metrics: demand external benchmarks, reproducible evaluations, third‑party audits and clear service‑level guarantees. Use skeptical frameworks that require evidence of generalisation, safety testing and documented human oversight.
-
What governance steps should I prioritise this quarter?
Quarter 1 actions: legal and procurement must roll out mandatory AI impact assessments for all new RFPs. Operations should require vendor carbon and water disclosures. Risk should schedule a tabletop exercise that tests both catastrophic and regulatory/PR scenarios within 60 days.
-
Are environmental and community impacts material to corporate strategy?
Yes. Hyperscale datacenters affect local utilities, water supply and social licence to operate. Incorporate infrastructure risk into site selection and vendor contracts, and require disclosure of energy and water footprints from suppliers.
A pragmatic reading plan and an action checklist
- Assign reading leads. Sponsor: one C‑level owner per theme, CIO for industry/reporting, GC for policy/ethics, Head of Sustainability for infrastructure. Deliverable: a one‑page memo with three strategic implications within four weeks.
- Extract operational rules, not just narratives. From each book, pull one actionable procurement rule, for example “no facial‑recognition deployments without legal and community review” or “require carbon and water disclosures in vendor contracts, ” and add the rule to procurement checklists.
- Run two tabletops. Use a long‑term alignment text to design a catastrophic‑risk tabletop and a surveillance/inequality text to design a near‑term regulatory/PR scenario. Compare playbooks and codify overlapping mitigations.
Three immediate actions to start this month
- Procurement: within 30 days, require AI impact assessments for any new RFPs. Owner: GC + Head of Procurement.
- Risk: schedule a 60‑day tabletop that uses one extreme alignment scenario and one near‑term surveillance scenario. Owner: Chief Risk Officer.
- Board: circulate the three‑book C‑suite shortlist and request a 10‑minute briefing at the next board meeting on AI exposure. Owner: CEO/Head of Strategy.
Parting note
Reading across genres, including investigative reporting, contested polemics, environmental studies and fiction, gives leaders a practical advantage. These books are not a call to panic. They offer focused signals. Use them to harden governance, reveal blind spots in procurement and infrastructure, and build measurable responses that keep your company resilient regardless of which headline trend dominates next.