🎆NYE Special: The Stories that Made 2025

When we launched in August, all we knew was that there was a gap. Political and policy professionals lacked information tailored to their work on Artificial Intelligence and some of the most consequential developments of our lives. From that simple idea, The Impact has grown in ways we could not have predicted.

In 2025, U.S. AI policy snapped from “cautious guardrails” to “full throttle” with Washington using AI money, rules, and infrastructure to reward speed and punish restraint. The Trump AI Action Plan and new OMB memos cast AI as a growth engine and culture‑war weapon: fast‑tracking data centers and federal adoption, favoring open‑source and “Buy American,” and threatening to starve states that insist on tougher protections or “woke” standards. Meanwhile, the public moved the other way, more anxious than optimistic, demanding testing, audits, and deepfake bans, and bipartisan in wanting someone to be in charge but not trusting either Big Tech or government alone. Taken together, 2025 revealed a quiet struggle over whose fears count more: voters worried about jobs, fraud, and fairness, or political leaders betting that visible AI expansion, especially in health care, benefits, and security, will pay off before the harms do.

For political and policy professionals, 2026 will test whether leaders can turn that gap into leverage: helping candidates to pick a side on deepfakes, state authority, and public‑sector AI tools. And answering the real question hanging over next year’s fights: when AI hits home harder for voters, will they decide they were served or sold out?

Takeaway

The White House released an AI Action Plan and three executive orders to drive innovation, data center buildout, and exports—and to block “woke AI” in federal use. The plan updates NIST’s AI Risk Management Framework, creates a DHS-led AI Information Sharing and Analysis Center, speeds federal permitting for data centers, and may limit funds to states with “burdensome” AI rules. It also revokes Biden’s January 2025 AI infrastructure order and sets up a program to export full‑stack U.S. AI tech.

Why it matters

The plan shifts federal AI policy from “guardrails first” to “growth first,” giving agencies cover to roll back Biden-era investigations and state-level rules as “burdensome.” That opens space for rapid AI rollout—especially in federal services and defense—while picking a partisan fight over “woke AI” that will shape what tools governments and vendors can use.

Takeaway

OMB issued two AI memos (M-25-21 and M-25-22) that replace Biden-era guidance and speed AI use across federal agencies. They set rules for “high-impact” AI, require each agency to name a Chief AI Officer, and push “buy American,” with new acquisition terms starting this fall. Expect tighter limits on vendors using agency data to train models without explicit consent.

Why it matters

These memos clear the runway for rapid AI expansion across federal agencies while loosening the Biden-era focus on civil rights language. That shift could normalize “high‑impact” AI in benefits, enforcement, and infrastructure decisions before guardrails are stress‑tested—setting up future fights over bias, transparency, and who controls government data.

Takeaway

Amid the shutdown, Paul Ryan urged Congress to tighten Medicaid rules, shift Medicare to private‑plan choice with capped subsidies, and fund state programs that cover the sickest patients. He called AI a cost fix and warned against strict rules, and said vague laws let presidents set health policy by executive order.

Why it matters

Ryan is packaging familiar GOP goals—shrinking Medicaid, raising retirement ages, shifting costs to “choice” and private plans—as technocratic fixes and AI optimism. For campaigns, this frames future fights: who pays for aging and innovation, and whether AI justifies cutting public coverage under the banner of “sustainability.”

Takeaway

GSA moved to make Elon Musk’s Grok AI available across federal agencies even though it has produced racist, antisemitic, and false content. Public Citizen and other groups asked OMB to stop the rollout, arguing it violates Executive Order 14319 that requires federal AI to be accurate and ideologically neutral. A White House science adviser also said Grok’s behavior breaks the administration’s own standard.

Why it matters

It signals that “AI safety” rules can be bent when they benefit powerful vendors and aligned ideologies—normalizing partisan, error-prone tools inside core government workflows. If Grok stands, it sets a precedent: future models won’t be judged by neutrality or truth, but by whose interests they serve.

Takeaway

A University of Maryland survey finds broad bipartisan support for government rules on AI. Majorities in both parties back testing AI before use, auditing systems already in use, banning deepfake political ads, and pursuing a treaty to ban autonomous weapons. The Trump AI Action Plan could still curb stricter state rules by tying federal funds to compliance, setting up a clash.

Why it matters

Public backing for AI rules undercuts GOP-style “hands off” arguments and gives cover to states and Hill offices to push harder—especially on deepfakes and audits. The fight now is whether Washington preempts stricter state laws and who gets blamed when AI harms show up in hiring, health care, or elections.

Takeaway

A review of 218 AI opinion studies finds people in the U.S. and U.K. are more concerned than optimistic about AI. Voters fear economy‑wide job loss more than personal job loss and want rules, but they don’t trust tech firms or government to do it alone. The authors launch AI SHARE to track attitudes and urge standardized, long‑term polling to guide policy.

Why it matters

This database shows AI is already a “live” political issue: voters are more worried than excited, fear job losses for society more than for themselves, and want strong rules but don’t trust either tech or government to write them alone. For campaigns, that’s a signal—and an opening—before narratives harden.

Takeaway

The Trump Administration released America’s AI Action Plan, prioritizing deregulation and tying AI money to states with fewer rules. The plan pushes open-source tools, speeds permits for data centers and chip plants, and tightens export rules while funding AI job training. Expect winners in low-regulation states and new risks in compliance and supply chains, especially around deepfakes and chips.

Why it matters

This plan weaponizes federal AI money to punish blue states and normalize deregulation, shifting power from safety-focused lawmakers to industry-aligned agencies. By preferring open-source and faster buildouts, it trades oversight for speed—raising the odds that harmful AI, synthetic media, and chip shortages hit voters before Congress can react.