The rise of artificial intelligence, and particularly of generative AI, is driving an unprecedented surge in demand for computing power.
It’s a spike that’s reminiscent of one we’ve seen before. In the early 20th century, the growth in electricity consumption outpaced the capacity of the existing grid. Today, data centers — energy-intensive hubs powering AI models — have become the modern-day “power plants,” and the strain they place on the power grid is comparable to the challenges faced by that early electricity infrastructure.
According to a report from the Electric Power Research Institute, data centers could consume up to 9.1% of U.S. electricity generation annually by 2030, up from an estimated 3-4% today. These soaring power needs could hinder the rapid adoption of AI — particularly in key markets like Northern Virginia, where the lead time to power new data centers can exceed three years.
The energy consumption at a data center has effectively become the bottleneck for AI advancement, blurring the lines between an energy utility and tech companies.
Both the electrification process and AI’s rise are linked to significant environmental challenges as the energy consumption of data centers grows. AI companies are beginning to invest in renewable energy sources and energy efficiency measures to mitigate emissions, including behind-the-meter solutions such as on-site renewable energy and nuclear power specifically for data centers. (That said, in many cases fossil fuels are still the fall-back.)
It may seem like an insurmountable problem, but there’s hope. The evolution of the electricity industry into a model of regulated monopolies offers a potential roadmap for the AI sector.
The potential for Public AI
In the early days of electricity, the high cost of building and maintaining infrastructure led to the creation of monopolies that were eventually regulated to ensure public access. Similarly, large tech companies currently dominate AI development, controlling significant portions of compute resources, data, and algorithms. This concentration of power raises concerns about equitable access and the risk of monopolization, particularly as AI becomes central to industries and everyday life.
The lack of a public AI ecosystem, experts warn, limits consumer choice and increases the risk of misinformation from AI-generated content. Just as electricity became a public utility with regulated access to ensure equitable distribution, there are growing calls for a “Public AI” option to prevent monopolization of AI resources.
Public AI envisions governments playing a role in owning and operating key elements of AI infrastructure, including data centers and AI model development. The Mozilla Public AI Report, released in September, emphasizes the need for publicly driven AI to foster transparency, inclusivity, and accountability. Mozilla — which was formed more than two decades ago as the explosion of the Internet set off similar concerns about regulation and access — advocates for the creation of public AI infrastructure centered on community-driven innovation and solutions tailored to underserved populations.
This public AI ecosystem would allow governments to ensure artificial intelligence resources, particularly compute power, are distributed more fairly. More than $850 million has already been invested in public AI labs, primarily nonprofits developing open-source AI models, showcasing the potential for this model to coexist with private development.
Challenges of directly translating the utility model
However, directly replicating the utility model for AI presents several challenges.
AI technology is far more complex and dynamic than electricity generation and distribution. Managing AI infrastructure requires expertise in intricate algorithms, massive datasets, and rapidly evolving hardware and software. Moreover, the artificial intelligence industry is characterized by constant innovation, and a heavily regulated utility model may stifle the kind of rapid development seen in the field today.
Also, the high cost of electricity infrastructure naturally led to monopolies in the energy sector, but the AI sector still has room for competition from smaller players. A monopolistic model for AI could inhibit innovation from startups and smaller developers who are driving some of the most groundbreaking work in the space.
But if a direct replication of the electric utility model for AI may not be feasible, there are alternative pathways that could offer solutions for the future of AI infrastructure:
- Hybrid models: A hybrid model could combine public and private ownership and operation of AI infrastructure. Governments could invest in public infrastructure while fostering competition among private developers, ensuring both equitable access and innovation.
- Regulation and standards: Governments will need to develop clear regulations and standards for AI development, deployment, and data usage. This could include guidelines for data privacy, ethical algorithms, and transparency in AI decision-making processes. The European Commission’s AI Act sets a precedent for regulating AI by focusing on safety, fundamental rights, and cybersecurity.
- Workforce development: Governments and industry must also invest in education and training to build a workforce capable of managing AI systems and driving future innovation.
Investment in Public AI
Governments also could play a significant role in funding research, developing open-source AI tools, and supporting public AI initiatives. By doing so, they can reduce reliance on a few large tech companies and create a more inclusive ecosystem.
The OECD AI Policy Observatory emphasizes the role of public investment in open-source tools and datasets — particularly for fostering an environment where artificial intelligence can be developed in a way free from harmful bias and more representative of the public’s needs. This would allow smaller players, academic institutions, and civil society to contribute to AI’s growth without being overshadowed by private sector monopolies.
In addition, philanthropies like the Open Society Foundations and its partners have collectively committed more than $200 million to ensure AI advances in the public interest. Their initiatives focus on promoting responsible innovation, enhancing transparency and accountability, and empowering workers to navigate AI’s transformative effects across industries. The overarching goal is to shape AI governance that protects democracy, human rights, and labor rights while fostering equitable access to the benefits of AI technology.
These efforts illustrate the growing momentum behind public AI initiatives aiming to democratize access to AI resources and ensure innovation serves the broader public good. They echo the response to the electrification of society over a century ago, when utilities were charged with the responsibility of providing energy to every home that wanted it. Now, as then, equity and sustainability must be the priorities.
Raghu Madabushi is an investment director at National Grid Partners, investing in early-stage companies in the energy transition and enterprise software vertical. He has more than 20 years of experience with technology, capital markets, and IP/innovation. The opinions represented in this contributed article are solely those of the author, and do not reflect the views of Latitude Media or any of its staff.
