Article illustration 1

Douglas Rissing/Getty Images

As the Trump administration prepares to unveil its formal AI Action Plan on July 23, 2025, a clear pattern has emerged: a relentless push toward deregulation, military integration, and private sector dominance, often at the expense of safety protocols and public research. This trajectory, established through executive orders, agency restructuring, and high-stakes contracts, signals a fundamental reorientation of U.S. artificial intelligence strategy away from the guardrails emphasized under prior leadership.

The Foundational Shift: Reversing Course on Day One
The administration's stance crystallized immediately. On January 23, President Trump overturned President Biden's October 2023 AI executive order, replacing it with a directive focused singularly on sustaining "America's global AI dominance." Crucially, terms like "safety," "privacy," and "consumer" vanished from the policy lexicon. This rollback, coupled with the simultaneous launch of Project Stargate – a data center initiative with OpenAI and foreign investors – signaled a philosophy where progress was framed as inherently opposed to precaution. "The rollback communicated the Trump administration's willingness to overlook the potential dangers of AI," noted Peter Slattery of MIT's FutureTech team, warning of a potential "Chernobyl moment" eroding public trust.

Dismantling Safeguards and Silencing Research
The subsequent months saw systematic erosion of oversight structures:
* Gutting the Safety Institute: The U.S. AI Safety Institute (US AISI), a Biden-era creation, saw staff cuts targeting researchers. By June, it was rebranded as the "pro-innovation, pro-science" US Center for AI Standards and Innovation (CAISI). Commerce Secretary Howard Lutnick declared, "Innovators will no longer be limited by these standards," explicitly positioning the center against "censorship and regulations." Notably absent were requirements for safety testing transparency or red-teaming reporting.
* Research Funding Cuts: The Department of Government Efficiency (DOGE) slashed grants for university AI research and education programs, raising alarms about the long-term U.S. talent pipeline. This occurred paradoxically alongside executive orders promoting AI "upskilling" for workers and students, highlighting a disconnect between rhetoric and resource allocation.

Military Embrace and the Global Tech Race
A defining feature of the new approach is the deepening entanglement between leading AI firms and defense:
* $200 Million Defense Contracts: In mid-July, the Department of Defense solidified major contracts with Google, OpenAI, xAI, and Anthropic. This followed Anthropic's launch of Claude Gov and OpenAI's broader government initiative, accelerating the integration of cutting-edge AI into military systems.
* Lifting the Nvidia Chip Ban: In a stark reversal aimed at fueling U.S. R&D, the administration lifted restrictions on Nvidia selling advanced AI chips to China on July 14. The rationale, championed by Nvidia CEO Jensen Huang and AI Czar David Sacks, posits that Chinese revenue will fund development keeping America ahead – a high-stakes gamble in the geopolitical tech race.
* Infrastructure Investment: A $92 billion investment announced in Pennsylvania for "AI and energy initiatives" underscored the focus on physical infrastructure to support domestic AI growth.

The Looming Policy and Unanswered Questions
The upcoming AI Action Plan is expected to cement this deregulatory, competitiveness-focused doctrine. Key unresolved issues remain:
1. Worker Displacement: How will the policy address the accelerating trend of companies replacing human roles with AI?
2. Safety Accountability: With CAISI eschewing mandatory safety reporting, what mechanisms exist to prevent harmful deployments or systemic failures?
3. State vs. Federal Power: Will the administration attempt to preempt state-level AI regulations, as hinted during recent Senate tax bill debates?

The Trump administration's AI legacy, thus far, prioritizes unfettered corporate advancement and military application as the path to dominance. The critical question for developers, researchers, and citizens is whether this speed comes at an unsustainable cost to safety, equity, and the foundational research that fuels true long-term innovation. The AI Action Plan won't just outline policy; it will define the rules of engagement for an era where technology increasingly shapes global power.

Source: ZDNET analysis of White House actions, executive orders, agency announcements, and industry developments (January - July 2025)