AI Week in Review - 7/5/25

Public Sector

Author

Chris Kraft
July 05, 2025

The Details

Federal

New OMB guidance from the Trump administration keeps Biden-era AI use-case inventory requirements largely intact but streamlines categories and re-labels “rights- and safety-impacting” systems as “high-impact.”

More than 60 organizations pledged resources to AI education under President Trump’s Executive Order, committing funding, curricula, tools, and mentorship to expand K-12 AI literacy over the next four years.

The U.S. Senate voted 99-1 to strip a proposed decade-long ban on state AI regulations from President Trump’s tax-and-spending bill after broad bipartisan opposition.

An Atlantic Council report warns civilian AI regulations can inadvertently restrict military innovation—shaping tech markets, legal authority, and costs—so defense leaders must proactively shape emerging global rules.

Under a new CRADA, NOAA’s National Hurricane Center will test and integrate Google DeepMind’s near-real-time AI tropical-cyclone forecasts, aiming to enhance hurricane track and intensity predictions.

State / Local

NACo’s new report showcases diverse AI and generative-AI pilots—from chatbots to wildfire monitoring—demonstrating how U.S. counties boost services, planning, and efficiency across five thematic areas.

Louisville’s IT agency will leverage a $2 million budget boost to pilot AI projects, appoint a chief AI officer and build a four-person AI team to streamline city services.

Georgia Court of Appeals fined attorney Diana Lynch the maximum $2,500 after discovering over 70 percent of her cited cases were fabricated or irrelevant, apparently AI-generated.

Bellevue, WA’s Safer Signals pilot uses Archetype AI’s sensor-based “Newton” to adjust traffic lights and instantly detect incidents, advancing Vision Zero’s 2030 goal to eliminate pedestrian fatalities.

International

The Netherlands will invest €70 million—plus seek equal EU co-financing—to build an AI research plant in Groningen by 2027, reducing reliance on U.S. technology and strengthening European autonomy.

UK Technology Secretary Peter Kyle directs the Alan Turing Institute to pivot toward defense, national security, and AI capacity, becoming a “national security asset” with new staff and refocused research.

Canada’s plan to scale AI in government must embed departmental AI leadership, robust oversight, and eco-efficient compute to avoid costly missteps and safeguard ethics, accountability, and climate goals.

England’s NHS will deploy an AI-powered early-warning system, starting with maternity outcomes in November, to flag patient-safety risks before Care Quality Commission investigations.

Seven tech firms pitched AI-powered home surveillance, drug-sniffing “smell-detectors,” and data tools to UK justice ministers, aiming for pilots that toughen 24/7 offender monitoring and curb reoffending.
AI, drones target illegal mining in Uttar Pradesh

Uttar Pradesh deploys AI-enabled check gates, drones, satellite imagery, and GPS tracking to curb illegal mining, blacklisting 21,477 vehicles while boosting transparency, enforcement, and environmental protection.

New Zealand’s Government Chief Digital Officer has launched an AI Expert Advisory Panel to advise agencies on responsible, efficient AI adoption and support public-service modernization with independent expertise.

Finance minister nominee Koo Yun-cheol will spearhead President Lee Jae-myung’s AI-driven growth agenda, coordinating cross-government investments and policies to raise productivity and counter Korea’s aging-driven slowdown.

Paris hosts 1,500 political and tech leaders, led by Macron and Modi, to forge ethical AI governance and investment, including France’s €109 billion pledge and EU supercomputing initiative.

Everything Else

Amid ongoing struggles with the LLM Siri, Apple is considering enlisting OpenAI or Anthropic to make up ground.

Executives who consulted ChatGPT became more optimistic, overconfident, and ultimately less accurate in predicting Nvidia’s one-month stock price than peers who simply discussed forecasts with colleagues.

Expert forecasters say AI chatbots that can outperform PhD virologists at bioweapon troubleshooting may boost the yearly risk of a human-caused pandemic fivefold unless strict safeguards are imposed.

Executives from Ford, JPMorgan, Amazon and Anthropic now openly predict AI-driven automation will wipe out up to half of U.S. white-collar and entry-level jobs, fundamentally reshaping corporate workforces.