News That Caught Our Eye #82

Published on October 30, 2025

Summary

Rethink AI finds that the 1,600+ AI-related state bills introduced emphasize guardrails over strategy. New Jersey releases practical guidance for building tools to simplify permits and answer resident questions. The UK launched AI-powered drug safety programs to predict harmful interactions before they happen, and NYC installed AI cameras in parks to detect brush fires in real time. Yet risks remain stark—a Virginia candidate staged a deepfake debate with his opponent, while speculative "superintelligence" manifestos distract from urgent governance challenges like surveillance and corporate accountability. Read more in the News that Caught Our Eye on AI and Democracy.

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.

Smarter AI Procurement: Lessons, Pitfalls, and Tools for Public Sector Leaders – October 30, 2:00 PM ET

Solving Problems with the Public – November 10, 12:00 PM ET

Co-Creating with the Multi-Lingual Communities – November 12, 12:00 PM ET

What Gets Measured Gets Done: Unlocking City-Level Health Data for Smarter Decisions – November 13, 2:00 PM ET

AI for Governance

Re-thinking AI: How a Group of Civic Technologists Discovered the Power of AI to Rebuild Trust in Government

Neil Kleiman, Mai-Ling Garcia, and Eric Gordon on October 27, 2025 in Reboot Democracy Blog

After two years of interviews, pilot projects, and field research, the RethinkAI collaborative has released Making AI Work for the Public—a sweeping review of how U.S. cities and states are using AI. The report finds that while 1,600+ AI-related state bills have emerged since 2019, most emphasize guardrails over proactive public sector strategy. Meanwhile, civic technologists and Chief Information Officers are leading bottom-up innovations. The authors call for a new governance model—ALT: Adapt to anticipate needs, Listen to understand communities, and build Trust through two-way accountability—to ensure AI strengthens, rather than erodes, public legitimacy.

Read article

AI for Governance

9 Tips for Building GenAI Tools for Public Sector Professionals

Jessica Lax on October 17, 2025 in New Jersey Office of Innovation

At the Office of Innovation, we’ve been building small but impactful internal-facing generative artificial intelligence (genAI) tools that are saving staff 83-99% on certain administrative tasks, freeing them to work on more complex and impactful work. What is the most significant lesson in building with genAI? Building AI products that work well for public sector professionals requires solving problems that go far beyond writing good prompts. There are technical hurdles, including timeouts and data processing quirks, as well as infrastructure decisions that can make or break your project. Below, we outline some of our key takeaways that guide our AI builds.

Read article

AI for Governance

Harnessing AI While Supporting Democracy

Kelly Born on October 16, 2025 in Project Syndicate

AI is reshaping democratic institutions across five concentric spheres—from elections and government efficiency to public participation, information ecosystems, and systemic risks. Kelly Born outlines both promise and peril: AI tools can streamline FOIA responses or enable participatory platforms like Taiwan’s Pol.is, but also risk deepfakes, algorithmic disenfranchisement, and democratic erosion. The article calls for foundational safeguards—strong privacy protections, lifecycle transparency, limits on surveillance, public AI infrastructure, and new liability and redistribution frameworks—to preserve trust and equity in an AI-powered era.

Read article

Governing AI

Governing the Undefined: Why the Debate Over Superintelligence Misses the Point

Dane Gambrell on October 29, 2025 in Reboot Democracy Blog

Amid new calls to ban “superintelligent AI,” this essay argues that the hype around artificial general intelligence distracts from real, urgent governance challenges. Gambrell critiques the AI doomerism movement for its vague definitions, conflation of imaginary risks with present harms, and its strategic utility to Big Tech firms seeking to shape regulation in their favor. Rather than chase speculative threats, he argues, democratic institutions must focus on tangible issues like surveillance, labor rights, and corporate accountability, problems already reshaping governance today.

Read article

AI and Elections

A Fake Debate in Virginia Raises Real Questions About AI in Politics

Kate Seltzer on October 23, 2025 in Governing / The Virginian-Pilot

Republican lieutenant governor candidate John Reid staged a YouTube debate featuring himself and an AI-generated version of his Democratic opponent Ghazala Hashmi, using synthetic voiceovers and selectively sourced responses. While Reid claimed transparency, the deepfake stunt sparked backlash from critics who saw it as misleading and provocative. The episode highlights how AI is reshaping political communication, raising concerns about disclosure, disinformation, and public trust as campaigns increasingly experiment with generative tools.

Read article

AI and Problem Solving

UK Launches AI-Powered Drug Safety Program to Predict Harmful Interactions Before They Happen

Staff on October 22, 2025 in Medicines and Healthcare Products Regulatory Agency

The UK’s Medicines and Healthcare products Regulatory Agency (MHRA) is spearheading three new AI-driven initiatives to modernize drug regulation, including a flagship project using NHS data and machine learning to forecast dangerous drug interactions before they affect patients. Targeting cardiovascular medications, the system will analyze real-world data to detect patterns, validate findings in lab models, and improve prescribing safety, potentially cutting billions in NHS costs linked to adverse reactions.

Read article

AI and Public Safety

FDNY Installs AI-Powered Cameras to Detect Brush Fires in NYC Parks

Phil Corso on October 23, 2025 in Gothamist

Amid rising wildfire risks, the New York City Fire Department is deploying solar-powered, AI-equipped cameras in parks to detect smoke and flames in real time. The cameras—already installed in Van Cortlandt, Highbridge, and Marine Park—stream live video to FDNY command centers, providing early warnings even in remote locations. As Commissioner Robert Tucker noted, the technology enables faster responses to prevent fires from spreading, complementing broader wildfire preparedness measures across the city.

Read article

AI and International Relations

Can India’s balanced approach to AI regulation foster innovation without stifling progress?

Lazar Radic on October 29, 2025 in Economic Times Commentary

India’s Competition Commission is taking a cautious, market-informed approach to AI, opting for study over sweeping intervention. Its September 2025 report highlights potential competition benefits while identifying emerging risks like algorithmic collusion and cloud infrastructure barriers. Instead of adopting rigid presumptions like the EU’s Digital Markets Act, the Commission emphasizes voluntary transparency, effects-based enforcement, and lowering structural barriers to innovation. The strategy suggests that in fast-evolving AI markets, regulatory calm may be India’s greatest competitive edge.

Read article

AI Infrastructure

OpenAI to expand into UK Data hosting after major growth deal

Staff on October 23, 2025 in Ministry of Justice

The UK Ministry of Justice announced a new partnership with OpenAI that will allow British businesses to host data on UK-based servers for the first time. The deal is part of a broader push under the UK’s “Plan for Change” to expand AI adoption, boost productivity, and reinforce cyber resilience. OpenAI CEO Sam Altman cited a fourfold increase in UK users over the past year, while Deputy PM David Lammy emphasized public sector applications, including the launch of “Justice Transcribe,” an in-house AI tool projected to save 240,000 staff days annually. The move ties into broader AI infrastructure plans like Stargate UK, powered by NVIDIA chips, and the creation of regional AI Growth Zones.

Read article

AI and Public Engagement

Facilitation in the AI Era: A New Toolkit for Democratic Deliberation

Ian Beacock and Emily Saltz, et al. on October 29, 2025 in Jigsaw (Google)

This new report from Google’s Jigsaw team offers a practical framework for adapting in-person and online democratic deliberation to an AI-saturated information environment. Drawing from case studies across 22 countries and multiple deliberative processes, the report outlines how AI-generated content and deepfakes are reshaping trust and facilitation practices. It introduces a new facilitation rubric to help practitioners design inclusive, adaptive, and resilient deliberation formats, from citizen assemblies to town halls.

Read article

▲