A living collection of readings for anyone working in, or curious about, public sector technology, AI ethics, and civic tech. Some are papers, some are books, some are talks. None of them will tell you which AI tool to use. All of them will help you think more clearly about what we're building and for whom.
Ethics
Many of these readings from a Data Ethics class at the University of San Francisco, syllabus prepared by Dr Rachel Thomas. Still a great class, though now taught by someone else. Enrolls once or twice a year.
- What is ethics? — A foundational primer from Santa Clara University's Markkula Center. Defines ethics not as a set of rules but as a practice of questioning what it means to live well together. Start here.
- A framework for understanding sources of harm throughout the machine learning process — Suresh and Guttag (2019) map out where bias enters ML systems: from data collection through model deployment. Essential for anyone evaluating AI tools for government use.
- Do artifacts have politics? — Langdon Winner's classic 1980 essay arguing that technologies embody political choices. The example of Robert Moses designing bridges too low for buses still resonates.
- An ethical toolkit for engineering / design practice — Practical exercises from Santa Clara University for teams building technology. Includes stakeholder analysis, consequence scanning, and ethical risk assessment templates.
- Anthropological/Artificial Intelligence & the HAI — Ali Alkhatib's sharp critique of Stanford's Human-Centered AI institute and the limits of "human-centered" framing when it's defined by a narrow slice of humanity.
- The Incompatible Incentives of Private Sector AI — Hagendorff and Wezel (2019) examine why profit motives and ethical AI development are structurally in tension. Relevant for any public sector worker evaluating vendor claims.
- How Algorithms Can Learn to Discredit "the Media" — Guillaume Chaslot (former YouTube engineer) explains how recommendation algorithms can learn to amplify distrust in journalism, even without being explicitly designed to do so.
- The Role and Limits of Principles in AI Ethics — Mittelstadt (2019) argues that AI ethics principles alone are insufficient without the professional structures (licensing, accountability, enforcement) that make medical and legal ethics work.
- Better, Nicer, Clearer, Fairer — Greene, Hoffmann, and Stark (2019) offer a critical assessment of the ethical AI movement, questioning whether its emphasis on fairness metrics distracts from deeper structural issues.
- Transcript of Dr. Rumman Chowdhury's talk on Algorithmic Colonialism at IntersectTO 2019 — Chowdhury connects the history of colonialism to modern data extraction, arguing that AI systems often replicate colonial power dynamics in how they collect, label, and deploy data from marginalized communities.
- The Call for Trauma-Informed Design Research and Practice — Rachael Dietkus, Social Workers Who Design (2022). Makes the case that design research, especially with vulnerable populations, must account for the trauma that participants carry. Particularly relevant for government service design.
Civic Tech & Digital Government
Books for anyone building, buying, or thinking critically about technology in public institutions.
- Recoding America by Jennifer Pahlka (2023) — The founder of Code for America and former U.S. Deputy CTO diagnoses why government technology fails: not for lack of money or tech, but because of rigid, industrial-era culture that separates policy from implementation. Named one of NPR's Best Books of 2023.
- A Civic Technologist's Practice Guide by Cyd Harrell (2020) — The most practical book in this list. Harrell draws on years of government tech partnerships to explain how to build alliances, navigate bureaucracy, and do lasting work in public institutions. Short, honest, and essential.
- Hack Your Bureaucracy by Marina Nitze and Nick Sinai (2022) — Nitze (former CTO of the VA) and Sinai (former U.S. Deputy CTO) share 56 tactics for getting things done inside large organizations. Written for anyone who has ever felt stuck, not just technologists.
- Power to the Public by Tara Dawson McGuinness and Hana Schank (2021) — A blueprint for public interest technology: put users at the center, use data smartly, run small experiments before scaling. Drawing on experience from USDS and the White House, with an afterword by Anne-Marie Slaughter and Darren Walker.
- Platformland by Richard Pope (2024) — The first product manager for GOV.UK maps out where digital public services should go next: 30 design patterns, 10 strategic interventions, and a compelling argument for "seamful design" over the seamlessness that big tech sells. The most forward-looking book on digital government published in recent years.
Product Management
- Impact First Product Teams by Matt LeMay (2025) — LeMay (author of Product Management in Practice and Agile for Everybody) argues that it doesn't matter how well you're "doing product" if you can't show business impact. Diagnoses the "low-impact death spiral" where teams gravitate toward safe work that creates little value. Applicable well beyond the private sector.
AI & Society
Books that examine AI's impact on labor, migration, equality, and the environment. Recommended for context, not as endorsements.
- The AI Con by Emily M. Bender and Alex Hanna (2025) — A sharp, witty takedown of AI hype. Bender (linguist, University of Washington) and Hanna (DAIR Institute) reframe AI as automation rather than intelligence, returning the conversation to real questions: what's being automated, who benefits, and who is harmed.
- The Oxford Handbook of AI Governance edited by Bullock, Chen, Himmelreich, Hudson, Korinek, Young, and Zhang (2024) — 49 chapters across 9 sections covering politics, democracy, economics, and administration in light of AI. The reference text for the field, published by Oxford University Press.
- Automating Inequality by Virginia Eubanks (2018) — Investigates how data mining, policy algorithms, and predictive risk models in welfare, housing, and child protective services target and punish poor people. Eubanks calls it the "digital poorhouse." Winner of the Lillian Smith Book Award. Still the most important book on algorithms and public benefits.
- Artificial Unintelligence by Meredith Broussard (2018) — A case against "technochauvinism," the belief that technology is always the solution. Broussard (NYU, Alliance for Public Interest Technology) argues that social problems don't inevitably retreat before digital tools, and that understanding what computers can't do is as important as knowing what they can.
- More Than a Glitch by Meredith Broussard (2023) — Broussard's follow-up confronts race, gender, and ability bias in tech, arguing these aren't bugs to be patched but features of how systems are built. Includes investigation of a mortgage approval AI that was 48% more likely to deny borrowers of color.
- The Walls Have Eyes by Petra Molnar (2024) — A lawyer and anthropologist documents how AI-driven surveillance technologies — robot dogs, drones, automated towers — are used to police borders and control migration. Finalist for Canada's Governor General's Literary Award.
- Atlas of AI by Kate Crawford (2021) — Crawford traces AI's supply chain from lithium mines to data laboring and back, revealing how AI systems depend on the exploitation of natural resources and human labor. A political and physical geography of artificial intelligence.
Further Listening & Reading
- Government and Technology Reading List — A community-maintained, crowdsourced reading list for folks in civic tech and public interest technology.
- Tech Policy Press: The Year in Books — Annual roundup of the most important tech policy books, with author interviews and podcast episodes.
This page is a work in progress. Suggestions welcome.