Reboot Weekly: Can A.I. Be Good for Democracy?

Published on October 8, 2026

Summary

New Jersey’s AI tool reached 106,000 more eligible children for summer food benefits, matching records across programs so no family had to fill out a form. In The New York Times, Beth Simone Noveck draws on that case, California’s post-wildfire listening tool, and the UK’s Consult platform that sorted 50,000 public comments in two hours to argue that AI can repair broken institutions and restore public trust. Also this week: Anirudh Dinesh traces today’s AI hype back to the 1920s invention of “halitosis,” Trump launches America.gov as a federal AI front door, and Newsom signs seven of nine union-backed AI bills.

Upcoming InnovateUS Workshops

InnovateUS delivers no-cost, at-your-own-pace, and live learning on data, digital, innovation, and AI skills. Designed for civic and public sector, programs are free and open to all.

TUESDAY, OCTOBER 13 · 12:00 PM ET: Prediction Isn’t Judgment — What Predictive AI Can and Cannot Do
With Aleksandra Korolova (Princeton) and Manoel Horta Ribeiro, moderated by Mihir Kshirsagar (Princeton). Predictive systems are increasingly used to forecast risks, allocate resources, and support public-sector decision-making — but prediction is not the same as judgment. This session explores how predictive AI actually works, what it can and cannot do, and the stakes for government decisions. Part of the Good, the Bad, and the Ugly of Predictive AI series.
[ Register for free → ]

Also coming up:

October 8, 2:00 PM ET — The Basics and the Breakthrough Opportunities (AI in Public Health series)
October 14, 2:00 PM ET — AI in Practice: How Public-Sector Lawyers Are Using AI (Using AI in Public Sector Legal Practice series)
October 15, 2:00 PM ET — Health Information in the AI Era: Combating Misinformation and Reaching the Public (AI in Public Health series)

[ See all workshops → ]

AI for Governance

Opinion | How A.I. Can Be Good for Democracy

Beth Simone Noveck on October 6, 2026 in The New York Times

Noveck argues AI can repair broken institutions and restore public trust in government. Drawing on New Jersey’s Office of Innovation (AI extended summer food benefits to 106,000 more children without a single form), California’s post-wildfire recovery (AI parsed 1,300+ survivor responses, now permanent under Newsom), and the UK’s Consult tool (50,000 public comments sorted in two hours), she proposes three investments: workforce AI training, secure access to multiple models, and putting AI behind every public servant.

Read article

AI for Governance

14 Big Ideas for the Next Era of Government

Zach Patton on September 30, 2026 in Government Technology

GovTech asked public-sector thinkers for one bold idea to fundamentally improve how government works. Neil Kleiman (Professor of Practice, Northeastern; Senior Fellow, Burnes Center for Social Change) argues the most important innovation of the next decade won’t be another chatbot or pilot — it will be the rise of communities of practice like the GovAI Coalition, Center for Civic Futures, RethinkAI, and InnovateUS, which are becoming the new civic infrastructure through which AI adoption spreads across American government.

Read article

Governing AI

The Halitosis of Our Times

Anirudh Dinesh on September 29, 2026 in Reboot Democracy

In the 1920s, Listerine sold “halitosis” as a hidden condition — sales rose from $115,000 to over $8 million by 1929. Dinesh argues AI companies are using the same playbook: describing software in sci-fi language (“agents go rogue,” “escape,” “scheme”) turns human and engineering failures into something that sounds beyond our understanding. On the July Hugging Face incident: OpenAI disabled cybersecurity safeguards during testing; the LLMs found credentials in an exposed dataset. The Bulletin of the Atomic Scientists breakdown shows human failures, not emergent machine agency.

Read article

Governing AI

Trump Launches AI-Powered Government Website America.gov

Associated Press on September 29, 2026 in PBS News

The Trump administration launched America.gov as a single AI-powered front door for federal services — from passport renewals to Social Security card replacements. Built by DOGE alumni (lead engineer Edward Coristine) and running on Google Gemini and Elon Musk’s Grok, the project was led by US Chief Design Officer Joe Gebbia (Airbnb co-founder). Trump signed an executive order at launch designating it “the single point of entry for Americans to access covered services online,” integrating Login.gov. Follows the GSA’s OneGov initiative and the USAi multi-model program.

Read article

AI and Education

Project Spotlight: Co-Designing an AI Tool with Communities

Joanna French on September 27, 2026 in Reboot Democracy

California’s AI-EP project made parents co-designers of an AI tool that makes their children’s education plans easier to understand. More than 200 families joined the first pilot and came away better able to advocate for their kids. It shows that who you choose to include shapes who the tool serves.

Read article

AI and Labor

On AI, Newsom Gives Labor Most of What It Demanded

Khari Johnson on October 1, 2026 in CalMatters

Governor Newsom signed seven of nine union-backed AI bills, including limits on AI in terminating or disciplining workers, requirements that healthcare professionals have final say in patient-care decisions, bans on employer AI-powered surveillance in bathrooms and emotional-state predictions, and mandates to disclose when layoffs are driven by AI. He vetoed two: healthcare worker retaliation protections and limits on AI in mental health settings. AFL-CIO called it “some of the strongest state legislation in the country.”

Read article

AI and Labor

Reclaiming Worker Power in the Age of AI

Sam Levine on October 2, 2026 in The Nation

NYC Commissioner Sam Levine draws parallels between the 1902 Lower East Side kosher meat boycott and today’s challenges of algorithmic pricing and personalized wages. He outlines a three-part playbook for responding at the city level: enforcing the law against AI-driven unfair practices, defending the shared economy, and investing in collective institutions that give workers organizing power in an era when every price and wage is personalized.

Read article

AI and Public Engagement

How AI Is Helping to Write a New Constitution in the U.S. Virgin Islands

Nathan E. Sanders on September 19, 2026 in Democracy Renovator / Rewiring Democracy Now

The U.S. Virgin Islands has convened its sixth constitutional convention — delegate Imani Daniel calls it “the first digital constitutional convention anywhere in the world.” AI is being used as an assistive tool in structuring a new government, framed by Daniel as a matter of self-determination for a territory with no voting representation in Congress, the presidency, or the UN. Part five in Sanders and Bruce Schneier’s series on pro-democracy technologies from their book Rewiring Democracy.

Read article

AI Infrastructure

Could AI Help Communities Navigate Complex Challenges Like the US Data Center Boom?

Jacob Taylor, Scott E. Page, Kershlin Krishna, Adam Aley, Vivek Ramakrishnan, Sarah Mills, Joshua Becker, Eirini Malliaraki, Fahd Husain, and Sanjay Purohit on September 30, 2026 in Brookings (17 Rooms project)

A 17 Rooms working group spanning AI, behavioral science, policymaking, and community organizing asks whether the same AI driving the data center boom could help communities hosting it make better shared decisions. Using a simulated Michigan county scenario, they workshopped design principles for community-led, AI-supported problem-solving — directly responding to the pattern Jasmine Sun named as “big companies descend on a small town and run roughshod over small-d democracy.”

Read article

▲