On June 25, 2025, Congress held a pivotal hearing regarding the United States’ approach to artificial intelligence, marking a significant shift in its stance on advanced AI developments.
Led by the Select Committee on the Strategic Competition Between the United States and the Chinese Communist Party, members from both parties voiced their concerns about the implications of artificial general intelligence (AGI) and artificial superintelligence (ASI), emphasizing the urgent need to address potential risks.
Representative Jill Tokuda of Hawaii expressed apprehension about a nation-state losing control over AI capabilities, warning of the rise of an independent AGI or ASI actor that could pose threats to global security.
Representative Nathaniel Moran of Texas predicted that AI systems would soon be able to conduct their own research and development, prompting concerns about the consequences associated with such advancements.
Representative Dusty Johnson of South Dakota highlighted the urgency surrounding the issue, stating, “Anybody who doesn’t feel urgency around this issue is not paying attention.”
As this historic hearing unfolded, a separate legislative initiative—the One Big Beautiful Bill—was progressing through Congress.
This extensive legislation aimed to cut taxes, ramp up military and border spending, and significantly reduce various social programs.
Republican senators contributed to the bill by worsening some of the safety net cuts included in the House version, along with an unsuccessful attempt to impose a new tax on clean energy that could hinder the construction of energy-intensive data centers required by AI advancements.
The contrasting actions within Congress reveal a disconcerting reality: while some lawmakers are beginning to acknowledge the significance of AI development, others are proceeding with a series of initiatives that could undermine the United States’ competitive edge in the growing field of AI.
In my previous observations, I emphasized that the One Big Beautiful Bill does not exhibit the foresight that one would expect from policymakers serious about harnessing the potential of powerful AI.
Notably, a provision passed by the House aimed to strip broadband funding from states that regulate AI, indicating a belief among its authors that AI does not warrant the same level of regulation as previous technological breakthroughs like telephones and the internet.
The Senate did manage to eliminate this provision on Monday, but the rest of the bill carries significant implications for the future of AI in the United States.
The legislation’s cuts to safety net programs, including Medicaid and SNAP (food stamps), signal a disregard for the potential job losses that may stem from automation resulting from advancements in technology.
As AI progresses, the likelihood of widespread job displacement grows, rendering the insistence on work requirements for support programs especially misguided.
For instance, if a person loses their job due to the increasing prevalence of self-driving services from companies like Waymo, access to essential resources like health coverage and food assistance will become tenuous due to the proposed alterations to these programs.
A person displaced from their job may find that their state opts not to provide food stamps at all or imposes work requirements that make it even more challenging to regain stability in a job market increasingly impacted by AI advancements.
Turning attention to the energy implications of the Senate bill, significant changes could resonate throughout the AI sector.
The House bill proposed drastic reductions in support for energy sources such as nuclear and geothermal, vital for the continuous power necessary for data centers and AI model training.
While the Senate did mitigate some of these cuts, energy tax proposals still emerged, such as accelerated repeal of wind and solar credits.
In the House version, renewable energy projects needed to be operational by the end of 2028, with the Senate version demanding that the timeline be shortened to the end of 2027.
An initial form of the Senate bill even suggested taxing wind and solar farms that relied on Chinese components, a problematic directive given China’s dominant role in global supply chains for renewable energy.
Fortunately, this tax proposal was ultimately removed from the final version of the bill, yet the continuation of wind and solar credit repeal still poses a concern for the AI industry.
For one, the changes are likely to inflate electricity costs, directly affecting operational expenditures for AI training and development.
The Rhodium Group, in modeling an earlier variant of the bill, projected increased energy costs for industry by 4 to 6 percent annually, primarily due to greater reliance on fossil fuels.
The reduction in renewable energy incentives means that projects may shift from clean sources like wind and solar to potentially pricier fossil fuels, which does little to alleviate the costs businesses would incur.
Furthermore, renewable energy sources such as wind and solar are variable in nature, which complicates their integration as reliable power supplies for data centers.
However, coupling wind and solar with battery storage technologies offers a viable solution, allowing for more consistent energy delivery—something that major tech companies like Google have already begun to adopt.
Given the rapid pace of AI advancements, timely construction of data centers, along with the associated energy infrastructure, is critical.
While nuclear power can provide a steady source of clean electricity, the lengthy construction timelines of recent nuclear plants can extend to over a decade.
Conversely, technological solutions like solar and wind systems, paired with batteries, boast a much faster implementation timeline, allowing for quicker scalability and adaptability to demand.
Statistics reveal that in 2024, 93 percent of new power capacity in the United States was derived from solar, wind, or battery technologies, indicating a shift toward these faster deploying solutions.
Failure to facilitate swift and economical access to clean energy sources could hinder the growth of AI firms, pushing them toward reliance on slower to build, high-emission alternatives.
The supply chain for natural gas turbines is presently strained in the U.S., and any shifts in demand brought on by the tax bill’s energy provisions could exacerbate these shortages further.
While some critics might interpret legislative slowdowns on AI development as beneficial, potentially mitigating risks associated with rapid AI progress, this perspective neglects the broader implications.
Rather than stalling AI advancements, these policies may instead catalyze a migration of data center construction to countries that favor green energy solutions.
As the global race for AI supremacy intensifies, nations like China stand to gain significantly by advancing their own AI industries, while the U.S. falls behind.
Such outcomes would not only undermine the competitiveness of American companies like Google and OpenAI but simultaneously bolstered opportunities for Chinese counterparts such as DeepSeek and Huawei.
Ultimately, the repercussions of the One Big Beautiful Bill could lead to a diminished role for the U.S. in the AI landscape—an outcome policymakers should strive to avoid.
What Congress seems poised to enact isn’t so much an industrial policy as it is a self-imposed handicap, endangering the nation’s standing in an arena where strategic foresight is critical.
It remains puzzling that so many lawmakers would willingly support a bill that risks ceding AI leadership to global competitors, underscoring the urgent need for a more comprehensive, forward-thinking approach to legislative success in the AI era.
image source from：vox
