Why US voters are fighting Trump’s AI data center boom — DW News

[00:00:00] Speaker 1: We begin the day with a question for U.S. President Donald Trump. How can the U.S. win the global race for artificial intelligence if most of America is standing in the way? Trump remains unyielding in his support for AI and those data centers. But three out of four American voters say they would oppose the construction of an AI data center in their own community. And 56% say that AI and data centers were a major factor in how they voted in the midterm election this week. Just consider the state of Texas. On Tuesday, Democrat Gina Hinojosa defeated Republican Governor Greg Abbott. She campaigned on a promise to suspend approval of new data centers until lawmakers put protections in place. [00:00:48] Speaker 2: The Greg Abbott corruption tax is why data centers have descended upon Texas. We have more than any other state in America. Greg Abbott invited them in with the richest tax breaks in the whole country. They pay no sales tax for 20 years. [00:01:10] Speaker 1: Well, for Texans, AI data centers present dilemmas as basic as who gets the water in a drought. And even Greg Abbott, who once celebrated the AI boom in the Lone Star State, he suspended environmental permits while officials checked how much water and electricity AI projects would need. And joining me now from Ann Arbor, Michigan is Shobita Parthasarathy. She's a professor of public policy at the University of Michigan, where she directs the science, technology, and public policy program. Professor, it's good to have you on the program. I understand that back in August, you said that data centers might actually get Americans out to vote. Now, we've had Election Day. We have a governor-elect in Texas who has promised a freeze on data centers. What evidence, though, do you see that this issue helped her beat Greg Abbott? [00:02:10] Shobita Parthasarathy: Well, she certainly had a very strong position on the issue. But I also think that if you look around the country, what you saw was 41 ballot initiatives in different states that were focused on data centers. And all but one, one in North Dakota, actually moved to restrict data centers in some way. And those are from all different kinds of localities. And I think we really saw a significant opposition and resistance to data centers across the country. So it certainly was part of the story in Texas as well. [00:02:49] Speaker 1: You know, Abbott himself was willing to put permits on hold, permits for data centers on hold. So how much common ground is there between Republicans and Democrats once they have to decide what it would take to start construction again? [00:03:07] Shobita Parthasarathy: Well, I think there's definitely alignment between Republican and Democratic voters. And they are really the ones leading the way. I think that Republican office holders are definitely following behind. And it's a pretty complicated situation in the U.S. right now, of course, because the president has been very strongly in favor of data centers. And many Republican office holders, like Abbott, have had to switch their positions in order to be in line with the voters. So I think the question now really is, are they going to take what the voters say seriously or not? In Texas, in particular, it's quite difficult because, of course, the richest man in the world lives in Texas. He spent quite a bit of money to affect Texas politics. And there is quite a bit of resistance to him. That is Elon Musk. And, you know, Elon Musk has done a lot of development in Texas. There are already communities that have felt the effects of, for example, SpaceX. And they are particularly exercised about that. So the dynamics are complicated, I think, in Texas. On the one hand, people have seen the real harms, the drawbacks of data centers and other kinds of AI-related development. But on the other side, you know, there's this historical focus and alignment now. We've seen it at least over the last couple of years between the Republican Party and the AI industry. And so that may be an alignment that's difficult to break. [00:04:51] Speaker 1: Let me pick up on what the consumers are seeing now where they get their electricity bill. You've got a company that comes to town and says that it will pay for every kilowatt it uses. So how could the family next door still end up paying more for power? [00:05:11] Shobita Parthasarathy: Well, what ends up happening, of course, is that overall there is much more energy use. So the prices end up going, there is a scarcity that's created and the costs go up. But I also want to make the point that what we're seeing here is that it is not just about energy costs. It is about all of the variety of different impacts of data centers, whether we're talking about noise pollution, whether we're talking about water scarcity. But also, as I suggested before, it's also about frustration with big tech and the power of big tech and the fact that they, you know, many citizens feel like they don't have power. And unfortunately, in a lot of these cases, when these companies have data center operators and big tech companies have come into municipalities, they've used some questionable tactics in order to push these data centers like nondisclosure agreements, which places additional burdens on these cities in terms of, you know, what does democratic governance look like? So I think it's difficult to look at the data center issue without those bigger, bigger concerns in mind. [00:06:27] Speaker 1: Yeah, I know that you've said that some of this anger is about inequality and also the perception of people being impotent in the face of government and business. Suppose the company that wants to build this data center, suppose the company covers the full cost of upgrading the grid and guarantees the town's water supply. How much of the opposition do you think would survive that deal? [00:06:54] Shobita Parthasarathy: I still think that opposition would continue. As I said, I really don't think it's just about individual costs. What's interesting to me about this groundswell of opposition is that it is really tied to a growing feeling of powerlessness. And we saw that yesterday, or I guess a couple of days ago now, there are polls that show significant opposition also to AI, majority opposition to AI. And we've been seeing that in poll after poll after poll. And there is a feeling of kind of really a lack of agency among citizens with the rollout of AI. And this is one of the few ways in which average citizens can actually push back against that. [00:07:41] Speaker 1: You know, I was struck by what we saw in Nottingham, New Hampshire this week. I think it was 9 in 10 voted to stop their town from accepting data center applications at least for a year. I mean, that is about as clear of a message as you're going to get in a democracy, isn't it? [00:08:05] Shobita Parthasarathy: Absolutely, absolutely. It's quite interesting. And, you know, I don't know that we generally think about such small towns, but it is really interesting to me that it is these small towns that in many ways are leading the way on this opposition. I only hope that elected leaders really listen. [00:08:24] Speaker 1: Do you think that people should have a more nuanced approach to this, that it can be okay to be against the proliferation of these data centers, and we don't want it in our backyard, but maybe that should not mean we're automatically against AI? I mean, do we need this type of dichotomy? [00:08:48] Shobita Parthasarathy: I mean, we are in a situation in the United States, unlike Europe, where we have no national level AI regulation. And so it's very difficult to say that, certainly very difficult to say that people's frustration and anger are unwarranted. And it is also understandable when people feel like, and they're absolutely right to feel, that in fact, there is no attention, no serious regulatory attention to AI. There is no, there has been no nuance in the conversation. And so what I hope that elected leaders take from this is that, yes, not only do they need to take, you know, policy on data centers seriously, they also need to take policy on AI seriously. In other parts of my work, what I find is that, you know, local governments are already adopting, you know, they're being bombarded by AI vendors with different products that claim to make governance better and more efficient. And often in these cases, they don't even know that they have to evaluate the AI, much less that they have to evaluate how to evaluate the risks and benefits of AI. And so we need real leadership and we do really need nuance, but that nuance needs to come from our policymakers and leading us in that nuanced conversation, as is happening in many other parts of the world. Unfortunately, we have not seen that in the United States, certainly in the last couple of years. [00:10:28] Speaker 1: Before we run out of time, let me, let me put the, the AI industries case to you. They would say that China is building and there are, you know, no bans or freezes there on data centers. If American communities keep turning these projects down, how serious is the risk that the U.S. loses its lead? And what happens when Washington says this is about national security? [00:10:53] Shobita Parthasarathy: Right. So I think a couple of things. The first and foremost, I think too frequently the AI industry relies on the lack of information, especially when it comes to China. In fact, China does have a pretty extensive regulation of AI. And, you know, we can quibble about what, what that looks like, but unlike the United States, China does regulate AI. The second thing that I want to note is that we tend to, and we certainly see this in the AI industry pose innovation and regulation as though they are opposite to one another. Yeah. They are not, um, guardrails and policies can help to shape innovation in ways that maximize social benefit. Unfortunately, the world that we have now is one in which the AI industry is calling the shots. Publics have absolutely no voice in this conversation. And so the kinds of AI that are being developed aren't necessarily in line with what publics want or what they need. And so I actually think if we have some regulatory infrastructure, it can help to ensure that we have innovation that's beneficial and in that way we can all benefit. And certainly those are the kinds of technologies that are likely to not only be more marketable, they're likely to lift all boats and, you know, help the country overall. [00:12:20] Speaker 1: Professor Parthasarathy at the University of Michigan Professor, good talking with you. Um, a lot of what you're saying is familiar to European ears. We'll see what happens, um, in the United States. Thank you. [00:12:37] Speaker 2: Thank you.