AI Week in Review - 12/20/25

Public Sector

Author

Chris Kraft
December 20, 2025

The Details

Federal

The Genesis Mission — launched via Executive Order on November 24, 2025 — establishes a new U.S. government initiative to build an integrated AI‑driven research platform under the Department of Energy.

New USPTO AI guidance confirms only natural persons qualify as inventors, even when AI significantly contributes to an invention’s development.

Congress will question tech firms after reports that Chinese-linked hackers used Anthropic’s AI in a sophisticated, largely automated cyberattack on U.S. infrastructure.

HHS launched a comprehensive AI strategy to modernize agency operations, improve public health outcomes, and boost efficiency through department-wide collaboration and responsible innovation.

President Trump’s December 11, 2025 Executive Order directs federal actions to preempt and challenge state AI laws, creating a Litigation Task Force and funding conditions to build a national AI policy framework.

OMB’s M‑26‑04 mandates that federal agencies require “unbiased AI” — truth‑seeking and ideologically neutral LLMs — in procurements, updating contracts and policies by March 2026 to build trustworthy AI use.

War Department Secretary Pete Hegseth launched GenAI.mil, a secure generative AI platform powered by Google’s Gemini for Government, giving all military personnel, civilians, and contractors access to AI tools to boost efficiency and operational tasks.

Pentagon CTO Emil Michael outlined plans under the Trump administration to rapidly deploy commercial AI tools across the Defense Department, equipping millions of users with capabilities tailored to operational and in‑office needs.

The 2025 President’s Management Agenda prioritizes using AI to streamline federal services, reduce duplication, and modernize government operations through secure, efficient, and data-driven technologies.

OpenAI and the U.S. Department of Energy signed a memorandum of understanding to deepen collaboration on AI research, deploying frontier models with national labs to accelerate scientific discovery under the Genesis Mission.

State / Local

Washington’s proposed House Bill 1622 would require public‑sector employers to negotiate with unions before deploying AI technology that affects wages, hours or working conditions.

36 state attorneys general urge Congress to reject federal proposals that would preempt state authority to regulate AI and protect public safety.

Tai Phan has been named the first Chief AI & Technology Officer of State of Oklahoma — tasked with leading AI adoption, setting ethical standards, and modernizing state government operations.

The NYC Council unanimously passed the GUARD Act, establishing an independent Office of Algorithmic Accountability — mandating transparency, fairness standards, and a public registry of all AI tools used by city agencies.

New York Gov. Kathy Hochul signed the RAISE Act, imposing broad safety rules and incident‑reporting requirements on advanced AI models, aligning with California’s framework amid lagging federal regulation.

Seattle has named Lisa Qian—an AI and data science leader from LinkedIn, Airbnb, and Convoy—as its first City AI Officer to implement its responsible AI strategy across departments.

California Gov. Gavin Newsom launched new initiatives—including a California Innovation Council and AI partnerships with top experts—to accelerate responsible AI integration in state government and modernize services with tools like Poppy.

International

France is investigating Elon Musk’s AI chatbot Grok after it generated Holocaust denial content in French, raising legal and ethical concerns over AI-generated misinformation and historical revisionism.

The APS AI Plan 2025 charts a roadmap for the Australian Public Service to harness AI across agencies — built on three pillars (Trust; People; Tools) — to deliver services faster, build agency capabilities, and ensure safe, ethical AI deployment.

The Bar Council’s Nov 2025 guidance urges barristers to use generative AI cautiously, ensuring confidentiality, accuracy, and personal responsibility remain central in legal practice.

Two southern England police forces are piloting “Bobbi,” an AI chatbot to field non‑emergency public inquiries — aiming to relieve pressure on call centers while keeping 999/101 human‑staffed.

Japan’s national ministries plan to deploy AI tools to help overworked bureaucrats draft responses and streamline administrative tasks amid heavy workloads, boosting efficiency in government operations.

China is advancing “embodied AI”—intelligence embedded in robots, vehicles, and environments—as a core path toward artificial general intelligence (AGI), supported by policy, research hubs, and national strategy.

The UK’s AI Security Institute’s Frontier AI Trends Report 2025 finds rapid capability growth in advanced AI—surpassing expert performance in science, cyber and coding while safeguards improve but vulnerabilities remain.

Everything Else

A China‑linked group designated GTG‑1002 used large‑language model agentic tools—chiefly Claude Code—to autonomously conduct 80‑90 % of a multi‑target cyber‑espionage campaign, from vulnerability discovery to data exfiltration.

Meta is testing an AI‑powered personalized morning briefing—internally called Project Luna—that analyzes users’ Facebook activity and external sources to deliver tailored daily updates, positioning it as a competitor to ChatGPT Pulse.

The report Anthropic’s “Estimating AI productivity gains from Claude conversations” (Nov 2025) quantifies how use of Claude across tasks and sectors appears to shorten work times — offering empirical evidence of AI-driven productivity gains.

An MIT study finds current AI could already replace 11.7% of U.S. jobs, mainly in admin, HR, logistics, and finance — totaling $1.2 trillion in wages.

HBR argues AI agents aren’t yet suited for customer-facing roles but can boost efficiency in structured, internal tasks like data handling, scheduling, and document processing.

ChatGPT’s growth has plateaued as rivals like Google’s Gemini surge ahead, raising doubts about OpenAI’s ability to sustain its lead and funding in the intensifying AI race.

A majority of enterprises now deploy AI agents for complex workflows, with 80% seeing ROI and plans to expand usage across engineering, reporting, customer support, and internal operations in 2026.

Enterprise spending on generative AI reached about $37 billion in 2025, more than triple 2024’s total, driven by broad adoption of AI applications and infrastructure across business functions.

Uber CEO Dara Khosrowshahi sees AI as delivering real productivity gains, enabling “superhuman” engineers, and accelerating autonomous vehicle ambitions through partnerships like Waymo.