Trump Rejects AI Guardrails as 'Sick Conspiracy' Amid Congressional Rift

What Happened

President Donald Trump opposed the implementation of regulatory guardrails for artificial intelligence on Monday, September 15, 2026. Writing on Truth Social, Trump characterized warnings from tech industry leaders regarding the need to slow AI development as a "sick conspiracy" against AI and data centers. He asserted that the only necessary guardrails are a "strong and smart president," claiming his administration already possesses the criminal and regulatory power to prevent companies from performing "bad things." Trump specifically targeted Dario Amodei of Anthropic, accusing him of pretending to be a "perfect little angel" while pursuing an IPO. Trump framed the issue as a binary geopolitical struggle, stating that whoever wins the AI race wins global dominance, and that only China benefits from U.S. regulation.

What the Evidence Establishes

The evidence establishes a significant fracture within the Republican leadership over AI governance. Senate Majority Leader John Thune explicitly broke from the president, stating that Congress can put guardrails around "consequential threats" without harming U.S. competitiveness. House Speaker Mike Johnson reported the creation of a bipartisan AI task force to research safety measures, though he opposed an emergency moratorium that would allow China to "overlap" the U.S. Other legislative efforts are already in motion: Senator Bernie Sanders and Rep. Greg Casar introduced the Ban Artificial Superintelligence Act on September 3, 2026, to prohibit AI humans cannot control. Additionally, Sen. John Kennedy announced he will introduce legislation requiring all AI models to include a "kill switch," a goal shared by Reps. Anna Paulina Luna and Ro Khanna.

Where the Accounts Conflict

Accounts conflict primarily on the motive behind industry warnings. House Majority Leader Steve Scalise claimed companies like Anthropic call for heavy regulation to "crush their competitors" and prevent new startups from entering the market. Conversely, a top researcher at Anthropic warned on Wednesday that there is a greater than 10% probability AI could "kill all humans" within the next decade. While Sen. Bernie Moreno dismissed these as "irrational fears" and "alarmist," pointing to Anthropic's IPO as evidence of hypocrisy, other lawmakers like Hakeem Jeffries argue the threats are immediate and require "decisive congressional action." Rep. Chip Roy expressed skepticism toward both sides, arguing that the "but China" justification used by tech firms is a "fear mechanism" rather than a viable policy prescription.

Context and Stakes

The stakes center on the concept of "technological singularity," described by Rep. Nancy Mace as the point when AI can autonomously create further AI. Mace argued that if China reaches singularity first, "all bets are off," as China would not honor any pause in development. Historically, the U.S. has balanced innovation with safety in sectors like the internet, a comparison made by Sen. Tommy Tuberville. The financial stakes are tied to the proliferation of data centers and the valuation of frontier model companies. Rep. Thomas Massie highlighted the political pressure within the MAGA movement, tweeting on August 31 that support for all data centers had become a "new directive from the leader," suggesting that AI infrastructure has become a litmus test for party loyalty.

What to Watch Next

Observers should monitor the progress of the "kill switch" legislation proposed by Sen. John Kennedy and Reps. Luna and Khanna to see if it gains traction despite Trump's opposition. A critical indicator will be the outcome of Speaker Mike Johnson's proposed meeting between the "seven or eight people" who control the major AI platforms and House leadership to find a "right balance." Additionally, any formal executive order from Trump regarding AI data centers or the restriction of "bad things" by AI developers will signal how he intends to exercise the "criminal and regulatory power" he cited. Finally, the movement of Anthropic toward its IPO will serve as a benchmark for whether the market views AI safety warnings as genuine risks or competitive maneuvers.

Bottom Line

President Trump has shifted the AI debate from a technical safety discussion to a nationalist geopolitical race, effectively attempting to preempt legislative guardrails with presidential discretion. While a faction of the GOP, including Steve Scalise and Nancy Mace, aligns with this "winner-take-all" approach against China, a substantial wing of the party—including John Thune and Mike Johnson—remains open to congressional oversight. The introduction of the Ban Artificial Superintelligence Act and proposed "kill switch" mandates indicate that a bipartisan consensus on existential risk exists independently of the executive branch. The conflict now rests on whether Congress will codify AI safety into law or defer to the president's claimed "High IQ" management of the technology.