he political backlash against artificial intelligence in the US is no longer a hypothetical risk.
Jefferies argues it is already gathering momentum across multiple fronts, creating a slow-burn threat to the industry’s expansion and, by extension, to companies betting heavily on AI infrastructure.
This is not a single skirmish but a broad shift in sentiment, visible in statehouses, courtrooms, opinion polls and even among the field’s leading scientists.
The most striking development is at the state level. In the absence of a federal framework, US states have surged ahead with their own AI laws, creating a patchwork of obligations that could slow deployment.
By mid-2025, 47 states had considered such legislation and more than 30 had enacted statutes covering everything from hiring bias to data-centre energy reporting.
California now requires safety frameworks and whistleblower protections for frontier AI developers, while New York’s RAISE Act imposes disclosure rules, safety obligations and penalties of up to $30 million for violations.
For developers, this is an increasingly complex compliance map to navigate. For investors, it marks the clearest institutional sign that the political tide is turning.
Attempts to move in the opposite direction have faltered. A proposed federal moratorium on state-level AI regulation, initially included in the One, Big, Beautiful Bill, collapsed under bipartisan criticism.
The Senate voted 99–1 to remove it, leaving the White House to explore an executive order that would lean on federal funding to curb state initiatives.
Jefferies notes that such a move faces constitutional hurdles, including limits on coercive spending conditions and a prohibition on forcing states to abandon their own laws. In other words, the fragmentation looks likely to persist.
Economic anxieties are adding fuel. In Congress, lawmakers are already probing whether AI is starting to displace workers. A bipartisan proposal would require large companies and federal agencies to report layoffs linked to AI, reflecting concerns that automation could push unemployment up by 10–20% within five years.
Early evidence remains thin, but Jefferies points to signs that entry-level roles in software development and customer service are being reshaped as firms embed more automation.
Industry pushback is also growing. As data-centre construction accelerates, heavy industrial users are increasingly wary of competition for grid capacity.
The report highlights the Louisiana Energy Users Group’s opposition to Meta’s new facility and notes that companies such as Diamondback and Devon Energy are now building their own power infrastructure to avoid congestion. It is a reminder that the AI build-out sits within a contested energy landscape.
Rising power prices are sharpening the political edge. Jefferies refers to retail energy data showing price increases in several states between 2019 and 2024, with AI-driven demand growth becoming part of election rhetoric in places such as Virginia and Georgia.
The recent election of Peter Hubbard to Georgia’s Public Service Commission is flagged as an early bellwether of public concern about power affordability.
Public sentiment is cooling too. Morning Consult polling on page 5 of the report shows declining favourability, trust and usage frequency for leading AI products since mid-2025.
Pew data, also cited in the note, reports that half of Americans are now more concerned than excited about AI in daily life, with younger adults particularly sceptical.
Even some of the sector’s pioneers are sounding the alarm. Jefferies highlights warnings from top AI scientists that rapid automation of AI research itself could propel the field far faster than governments can regulate.
Geoffrey Hinton’s caution that AI will worsen inequality by replacing workers underscores how the debate has moved beyond technical risk to broader social and economic questions.
For investors, the thread running through Jefferies’ analysis is clear. The political, regulatory and social foundations that underpinned the first wave of US AI enthusiasm are now shifting.
Whatever happens in Washington, the next phase of AI growth will unfold under far closer scrutiny than the last.