���Ҵ�ý�ƽ������

Since OpenAI's ChatGPT launched in November 2022, generative artificial intelligence, particularly large language models (LLMs), have become widely demanded across the globe, in what is termed the "AI Boom" (HISTORY.com Editors, 2025). As of July 2025, the widely popular ChatGPT received over 2.5 billion queries per day, with roughly 10% of the world's adult population using the platform (Adams, 2025). OpenAI represents only a part of the much broader AI market, alongside companies such as Google, Microsoft, Meta, and Anthropic. Taken together, the rise in artificial intelligence is disproportionate. By some estimates, AI-related hardware and software, including data centers, chip-making facilities, and new models and tools, accounted for around 1% of the United States' GDP growth in 2025 (Collyns, 2025).

AI's rapid growth represents a new political force generating conflicts over energy and water use. As governments try to grapple with climate change and enforce environmental policies, AI poses a fundamental challenge by introducing a resource-intensive industry that strains existing environmental policies. Governments face major tradeoffs between the economic benefits of AI and major localized environmental costs.

Perhaps the most evident and visible example comes in the form of data centers. Despite AI's digital nature, it requires vast amounts of physical infrastructure, particularly data centers to house large, complex algorithms. While data centers (places that store, process, and distribute digital data) are nothing new, as digital workloads increased during the 2010s, energy and water demand remained relatively constant due to technological advancements in cooling and hardware (Iman, 2025; Schmelzer, 2026). Unlike these stable trends, the disproportionate scale and rapid expansion of AI computing demands have surpassed the capabilities of current data center technology to manage land, energy, and water.

Hence, more data centers are needed to keep pace with technological advancements, and the United States, which houses a third of global data centers, is a major driving force (Iman, 2025). Yet, with data centers popping up, policymakers and residents are increasingly worried about their potential impacts on communities, such as higher utility rates and pollution.

For illustration, globally, data centers account for 1-2% of electricity use, about the energy usage of a country the size of Sweden or Argentina, and are "likely [to] reach 3%–4% by 2030" (Iman, 2025). In the United States, water used for data center cooling "is equivalent to the water consumption of a city of 3 million people" (Iman, 2025).

The rising energy and water demands are not just an abstract concept but are increasingly shaping local politics. Communities and policymakers nationwide are forced to balance tradeoffs between economic development and environmental stability.

In Colorado, a few companies have begun construction on data centers, prompting various responses from residents. Larimer and Logan Counties have enacted moratoriums to halt construction until state regulatory laws are enacted (Schmelzer, 2026). In 2026, Colorado lawmakers will vote on two bills, House Bill 26-1030 and Senate Bill 26-102, that propose radically different approaches to data centers. These bills will be another attempt to regulate this emerging industry.

HB 26-1030 proposes conditional tax breaks to data centers that meet certain requirements. In this way, the HB hopes to improve Colorado's competitiveness for AI data centers while limiting the environmental impacts and preventing higher household utility rates. To receive the bill's tax cuts (a 100% state sales and use tax exemption), the data center must invest $250 million in the center, create high-paying jobs (110% of the county's average wage), use waterless cooling technology, and transition to renewable energy. Proponents argue that HB will attract high-wage jobs and invest in communities while still protecting scarce water resources and energy rates. By some estimates, tax breaks would decrease state revenue by "$92.5 million in sales tax revenue in the first three years, assuming a total of 17 data centers would qualify for the tax breaks in that time period" (Schmelzer, 2026).

Alternatively, SB 26-102 proposes to enact high environmental standards and prohibits offering economic development rates to large data centers. Data centers will be required to report energy and water usage, and by 2031, all their annual energy consumption will be from renewables. The bill also prohibits utilities from offering economic development rates to large load data centers. Proponents say that data centers will pay the full incremental cost of infrastructure rather than passing higher rates onto individuals and small businesses. Further, it will allow Colorado to achieve or protect its current environmental goals that are threatened by data centers. Those opposed argue that this will drive data centers away from Colorado and question the overall bill structure. For example, requiring dry-cooling systems to save water, yet such cooling systems can lead to a 13% annual energy increase (Kapiloff et al., 2025).

While it is unlikely that either of these bills will pass in their current form, they do clearly illustrate competing interests between economic development and environmental protection. Yet, even if Colorado and other states attempt to drive data centers away, it is unlikely that these conflicts will remain regionally isolated. The increasing online nature of today's society and the rapid adoption of AI technologies have created a need for a place to house and process large amounts of digital data. Governments will likely continue to confront tradeoffs between attracting investment and managing environmental costs. Further, current environmental regulations will need to be updated to account for these new factors. Still, the politics surrounding AI infrastructure highlights a broader challenge to modern governance, one that has been highlighted since the introduction of the internet: how to regulate our evolving online world, whose material demands are often obscured by its very digital nature.

��

References

Adams, M. (2025, October 16). ChatGPT is processing 2.5 billion messages a day. Business Insider.

Brasch, S. (2026, January 21). Colorado lawmakers brace for their biggest battle yet over AI data centers. Colorado Public Radio.

Collyns, C. (2025, December 19). Tracking the AI Boom: Some Lessons From Economic History | Econofact. Econofact | Key Facts and Incisive Analysis to the National Debate on Economic and Social Policies.

HB26-1030 Data Center & Utility Modernization | Colorado General Assembly. (n.d.).

HISTORY.com Editors. (2025, November 25). ChatGPT is released to the public | HISTORY. HISTORY. ��

Iman, N. (2025). The Hidden Costs of Intelligence: Artificial Intelligence and Machine Learning Adoption and the Paradox of Exponential Digital Growth. Sustainability and Climate Change, 18(3), 225–241. ��

Kapiloff, D., Rogers, L., & Tellinghuisen, S. (2025). Data Center Impacts in the West: Policy Solutions for Water and Energy Use. Western Resource Advocates.