Trump’s 90-point AI Action Plan promises rapid innovation through deregulation, but critics warn of bias, job losses, and environmental costs.
In July 2025, the Trump administration released a 28-page blueprint, “Winning the Race: America’s AI Action Plan,” which reads like a modern-day gold-rush map. It outlines over 90 policy positions across multiple agencies, all with a single goal: to remove barriers to AI innovation. This deregulatory approach is the heart of the plan.
Why It Matters Now
With China, the EU, and private rivals all racing to lead in AI, the administration argues that streamlined approvals and clearer guidelines will help U.S. firms innovate faster. Critics counter that speed may come at the expense of environmental safeguards, worker training, and protections against bias.
Staking the Claims: Anatomy of a Deregulatory Plan
The AI Action Plan isn’t a single law. It’s a series of executive orders and policy mandates designed to remove regulations and accelerate AI deployment. Key elements include:
- Fast-Tracked Permitting: An executive order specifically expedites federal permits for data centers and semiconductor manufacturing under existing NEPA and FAST-41 processes. This directly responds to a major industry complaint about infrastructure build-out delays.
- AI Export Promotion: The Commerce and State Departments will partner with industry to export “secure, full-stack AI packages” to U.S. allies. This policy aims to build an American-led AI ecosystem abroad, free from foreign regulatory influence.
- “Woke” AI Guardrails Removed: New procurement rules will expunge DEI language from federal contracts, insisting that federally contracted AI must reflect “objective truth” without ideological bias. This is a clear move to deregulate the ethical and social guardrails placed on AI development.
Prospecting for Performance: Technical Leaps & Public Pulse
The administration’s deregulatory push coincides with rapid technological advancements. The plan aims to build on these successes by removing what it sees as unnecessary red tape.
- Medical Device Claims: The FDA cleared 221 AI-enabled medical devices in 2023, up from just 6 in 2015. This surge in regulatory confidence directly results from new policies that allow companies to test and deploy AI tools more quickly.
- Benchmark Breakthroughs: AI performance on major benchmarks saw dramatic leaps in 2024. The MMMU, GPQA, and SWE-bench test scores rose by 18.8, 48.9, and 71.7 percentage points, respectively. The plan argues that removing bureaucratic friction will accelerate this progress even further.
- Public Sentiment: This progress is met with public skepticism. A 2025 AI Index report found that only 38% of Americans believe AI will improve health and only 31% expect net job gains, a sentiment that echoes the wary attitude of a miner looking for fool’s gold.
Those gains, equivalent to finding gold flakes in untested soil, suggest that models learn faster. But breakthroughs on test benches don’t always match real-world reliability.
Data Centers: Growth and Impact
The new permit rules have unleashed a wave of data-center proposals:
- Energy Use: U.S. facilities consumed 176 terawatt-hours in 2023 (about 4.4% of national electricity) and could reach 12% by 2028.
- Emissions Toll: A Department of Energy survey of 2,100 centers found 105 million tonnes of CO₂ last year, more than half from fossil-fuel backup generators.
Faster approvals mean new investment dollars and sharper debates over rising energy demand and the environmental footprint of an AI boom.
Chips and Open Source: Who Benefits
Hardware and community code are twin engines of the AI economy:
- Semiconductor Exports: American chip sales hit $70.1 billion in 2024 (up 6.3%), driven by fabs in Texas and Oregon.
- Model Scans: Open-source security tools have analyzed 4.5 million AI models and flagged 350,000 potential biases or safety issues, proof that not every discovery is pure gold.
Eased export rules give chipmakers new markets, while looser sharing lets small labs, from university groups to bootstrapped startups, compete with hyperscale giants on the same playing field.
Jobs at Risk and Opportunity
No gold rush is without its claim jumpers and ghost towns:
- Automation Risk: A McKinsey study warns that 30% of U.S. work hours could be automated by 2030, triggering 12 million occupational shifts.
Commenting on the human cost of these changes, Anirudh Agarwal, Director at OutreachX, cautions, “Accelerating permits without investing in people is like staking gold claims with no plan to refine the ore.”
Claim Holders and Ghost Towns: Potential Winners & Losers
The deregulatory “gold rush” is creating clear winners and losers.
-
Winners:
-
-
- Chip Makers & Fab Operators: Can build new semiconductor “mines” under eased zoning regulations.
- Cloud Giants: Can erect hyperscale campuses with fewer permit delays.
- Open-Source Labs: Are designated as official prospectors, free to pan for new open-source models.
-
-
Losers:
-
- Front-Line Workers: Face shuttered roles without guaranteed retraining.
- Civil Rights Advocates: Warn that removing DEI guardrails may lead to biased or unsafe AI in critical services.
Civil Rights and Accountability Concerns
Several advocacy organizations have raised alarms about the broader impact of unfettered deregulation:
- ACLU: The plan undermines state authority by directing the Federal Communications Commission to review and potentially override state AI laws, while cutting off ‘AI-related’ federal funding to states that adopt robust protections,” Cody Venzke, Senior Policy Counsel with the American Civil Liberties Union.
- People’s AI Action Plan: Over 80 labor, civil-rights, and environmental groups released a rival blueprint, warning that unfettered deregulation caters to Big Tech, sidelines public interest, and undermines worker protections.
- State Protections: Critics note the federal plan overrides thoughtful local safeguards, stripping states of the right to prevent AI-driven bias in housing, healthcare, and law enforcement, and risks “unfettered abuse” of AI systems.
Mapping the Aftermath
Deregulation has opened the sluices for an AI gold rush, fueling boomtowns in tech hubs and reshaping local economies. Yet, as with every frontier rush, the real test comes when the veins run dry. Will communities that staked their claims emerge wealthier, or face the ghost-town fate of those left sifting yesterday’s tailings? As Congress, courts, and citizens weigh in, the question remains: in this 90-point gold rush, who finds riches and pays the toll?
