AI Data Center Boom Faces New Risks as Public Backlash and Policy Scrutiny Grow

Public opposition to large-scale data center construction is rising sharply, creating new challenges for the artificial intelligence (AI) infrastructure boom, according to Wolfe Research. The firm identified growing voter resistance and increasing government oversight as the two most significant near-term risks to AI-related capital spending.

Recent polling cited by Wolfe Research showed that 53% of respondents now blame data centers for rising electricity prices, nearly doubling from 28% just nine months ago. The issue has become the leading public concern over higher power bills, ranking ahead of oil and gas companies, utilities, aging grid infrastructure, the Trump administration, and geopolitical tensions.

The policy shift is already taking shape. New York Governor Kathy Hochul recently imposed a one-year moratorium on permits for new data centers with power demands of 50 megawatts or more through an executive order. Wolfe analysts noted that executive actions like New York’s could spread faster than legislative efforts, even though similar proposals in at least 15 states have largely failed, stalled, or been vetoed.

Despite the growing political pressure, Wolfe Research does not expect a nationwide wave of bans. Instead, data center investment is likely to shift toward states such as Texas, which continue to actively encourage AI infrastructure development.

Federal AI regulation is also evolving. Analysts said the Trump administration, initially viewed as largely hands-off, has become more involved as advanced AI models raise cybersecurity and national security concerns. The White House reportedly relied on early access to Anthropic’s Mythos model for cybersecurity assessments before imposing export restrictions on certain frontier AI models. OpenAI also agreed to initially limit access to its GPT-5.6 models to selected trusted partners at the government’s request before expanding availability.

In addition, President Donald Trump signed an executive order requiring certain frontier AI models to be shared with the U.S. government and designated partners 30 days before public release, with implementation guidelines expected by early August.

Wolfe Research also expects Washington to discourage the adoption of Chinese open-weight AI models over security concerns, a move that could benefit U.S.-based AI providers. While the firm believes current policy changes are unlikely to significantly slow AI capital spending in the near term, it warned that tighter regulations could become more restrictive as AI technology continues to advance.