AI: Sustainable and Strategic

The views expressed are those of the author and do not necessarily reflect the views of ASPA as an organization.

By Parisa Vinzant
August 29, 2025

The case for making AI innovation both sustainable and strategic has never been clearer or more pressing. To allow industry to drive AI innovation without sufficient policy direction and oversight is a risky gambit that introduces the likelihood of unintended negative consequences. Beyond the major issues discussed later in this article are emerging problems for public servants to address, including AI models’ extraordinary carbon footprint and the enormous drinking water and electricity usage at AI data centers. It will be costlier to mitigate these consequences later than address them at the start. Unfortunately, it is typically the public, not industry, that disproportionately bears such costs. A brief comparison of the two approaches taken by the United States (US) and the European Union (EU) to AI innovation, particularly as it relates to generative artificial intelligence (GenAI), proves instructive and reveals what the future may hold.

In the US, industry has largely been allowed to advance AI innovation without coordinated policy or strong oversight. By their nature, for-profit companies prioritize decisions that are in the best interests of the few over the long-term well-being of the many. The lightning-fast progress in GenAI development taking place in the US would be impossible not only without vast amounts of drinking water and electricity, but also the expansion of gas networks to meet rising demand. Of particular concern is the fact that drinking water is a finite resource, and a public good first and foremost. There are growing concerns that public officials have not always adequately considered their constituents’ interests when signing contracts with AI data centers. Yet, these officials have a responsibility to empower themselves with resources such as the operational water footprints for AI and the data center water consumption calculator to reduce any informational or power disadvantages when negotiating with AI data centers; in this way, the long-term well-being of their constituents can be better protected.

In contrast, through its AI continent action plan, the European Union is taking a leadership approach to shape, fund, and deploy “AI solutions that benefit society and the economy.” With well over €200 billion currently allocated, there are five key focus areas: building a large-scale AI computing infrastructure, increasing access to high-quality data, promoting AI in strategic sectors, strengthening AI skills and talents, and simplifying the implementation of the AI Act. The last point is especially important since the EU has an uneven track-record of rapid innovation deployment because it must act collaboratively and it tends to over-rely on regulation. Yet, the EU approach is intriguing because public sector values are at the core along with economic values. In one example, it appears the planned Mistral 1.4 GW power data center cluster outside Paris could exemplify sustainable solutions such as low-carbon energy generation. In general, sustainable innovation is already foundational in Europe and the issues of massive water and energy usage may be better addressed by AI innovators there.

Though the US maintains global leadership in AI-focused data center capacity, the EU continues to invest in and revise its long-term AI action plan. It’s too soon to count the EU out in the global AI innovation race, as its pragmatic approach may prove strategically wise when all considerations are factored into the ever-evolving AI innovation development picture.

It is crucial to grasp the role that inadequate, disjointed policy or the lack of regulatory controls played in accelerating GenAI development and data center growth in the US. These factors underscore a key weakness and a potential threat to the United States’ current dominance as well as global progress on AI innovation. The US government has essentially empowered the AI industry to such a degree that they may prove to be ungovernable.

Let’s consider this hypothetical scenario: what happens if, or when, artificial general intelligence (AGI) is achieved within this lax regulatory, vision-deficient US context? According to IBM, AGI is a hypothetical stage in the development of machine learning where AI systems can match or exceed the cognitive abilities of humans in any task. Those most closely involved with generative AI in the US are already racing ahead on the next monumental AI innovation toward AGI. While there is current debate as to whether AGI is even feasible, it is vital to be aware of what these AI researchers are doing with their abundance of funding.

Taking this scenario further, consider the warning by ex-OpenAI prodigy Leopold Aschenbrenner in his June 2024 piece that “as the race to AGI intensifies, the national security state will get involved.” Aschenbrenner advises deployment of a project like the Manhattan Project because “no startup can handle superintelligence.” The timing for this anticipated national defense intervention is estimated to take place around 2028. For Aschenbrenner, the importance of the national defense intervention is simple: “it will become clear that the development of AGI will fall in a category more like nukes than the internet.” But for the US government to intervene at a relatively late stage in AGI development and implement “a sane chain of command for superintelligence,” as advised by Aschenbrenner, presents a daunting challenge.

It is very difficult, if not impossible, to put the technology genie back inside bottle and contain it after it exists. This is why, in the long run, the EU’s approach to AI innovation may be more strategically sound, safe and sustainable.

---

Author: Parisa Vinzant, MPA, works as a private and public sector strategist. She provides coaching to ICMA members. She served as a technology/innovation commissioner in Long Beach, CA. Parisa applies an intersectional equity lens in her writing on topics such as technology, ethics, and democracy. Connect with her by email at [email protected] or on Bluesky @pvinzant.bsky.social. Signal contact by request.  

(3 votes, average: 5.00 out of 5)
Loading...