In an exclusive interview, prominent American technologist Bruce Schneier tells Cybernews that Elon Musk’s plans to automate the US federal government and use AI to fix fraud are “total cyber utopianism.”
“Psychoanalizing Elon Musk is beyond my pay grade. I don’t know what he’s doing. I wish I did. We all wish we did,” admits Schneier, a fellow at the Berkman Klein Center for Internet and Society at Harvard University and a lecturer in Public Policy at the Harvard Kennedy School.
He’s definitely worried, though. That’s because Musk’s Department of Government Efficiency (DOGE), probably the most important or, at least, most active agency in the Donald Trump administration, is running wild in Washington.
Schneier already wrote in Foreign Policy that DOGE is “hacking America” by unlawfully accessing ultra sensitive government systems.
These external operatives are doing their work in plain sight, he said, whereas foreign adversaries would typically spend years “attempting to penetrate systems such as these, using stealth to avoid being seen, and carefully hiding any tells or tracks.”
“Techno-fascism by chatbot”
However, even more bizarre plans, soaked in open contempt for federal bureaucracy, are being brewed, apparently.
The New York Times and other news organizations reported recently that Musk allies, pushed inside government agencies, are thinking about how to harness AI to identify cuts of both budget and people, and to detect waste, abuse, and fraud.
According to one official who spoke to The Washington Post, the ultimate goal is to use AI to replace “the human workforce with machines.”
Does it look a bit like a wet dream of a soulless libertarian tech visionary to whom humans are tools rather than, well, humans?
“Oh, God, yes. Totally cyber utopianism,” Schneier tells Cybernews, adding another description to Musk’s vision, also recently called “techno-fascism by chatbot” by Kyla Chayka in The New Yorker.
The prominent technologist is additionally worried about the possibility that the Trump administration might be seeking to automate democratic decision-making rather than simple paperwork.
Eryk Salvaggio wrote last week on Tech Policy Press that what DOGE is doing amounts to political offloading, “shifting the messy work of winning political debates to the false authority of machine analytics” and replacing government bureaucracy with technical infrastructure.
“The idea of using AI in government is not stupid in the main as long as you do it right. You have to think about what you’re doing and do it right,” Schneier says.
The Biden administration, for instance, disclosed more than 2,000 AI applications in development across the federal government.
The Federal Emergency Management Agency has started using AI to help perform damage assessment in disaster areas, and Centers for Medicare and Medicaid Services are using AI to look for fraudulent billing.
“But with Musk and DOGE, we can’t talk about them as if what they’re doing is normal because it’s not,” Schneier points out.
Sure, one could say that the best way to get rid of human error is to get rid of humans. But human expertise still seems important. Even bureaucrats improvise and use discretion, and computers are good at following instructions we give them, not improvising.
“It’s AI versus nothing now”
Schneier is now writing a book on AI and democracy where he analyzes the different ways AI can make bureaucracy more efficient. He sees a lot of positive use cases and believes that AI can definitely be useful in bureaucracy.
“For example, can AI audit everybody’s tax returns? US Republicans will never want that but can it? The AI doesn’t assess fines – it just highlights tax returns that a human should look at. That’s all it does, an initial call,” says Schneier.
He suggests looking at AI as a partner because humans and AI should work together: “Let’s have the AI tell the humans what to do. In the US, there’s a two year backlog for disability benefits, for example. AI could help, and we could make it only say yes.”
“But with Musk and DOGE, we can’t talk about them as if what they’re doing is normal because it’s not,”
Bruce Schneier
“All it could do would be to make the easy yeses – because, after all, most of these are easy yeses. Anything that’s hard, anything that’s a no, would kick to a human,” says Schneier.
“And there were hundreds of these ideas coming from the Biden administration.”
“This is not Musk going in, copying everything, and doing awful things. It’s just thinking about where we don’t have enough humans. Yes, in some of these cases, AI isn’t better than the human. The problem is that it’s not AI versus human – it’s AI versus nothing,” he adds.
“We’re unable to audit the tax returns we want to. AI misses stuff but not auditing at all misses even more things.”
Nicholas Carr wrote many years ago that “there is a conflict in many cases between the desire to maximize efficiency through automation and the desire to make sure that human skills, human talents, continue to be exercised, practiced, and expanded.”
So far, though, it doesn’t really look like humans are appreciated in the Elon universe. Schneier doesn’t do drama, though. He says: “They’re not trying to destroy the government. They’re pretending.”
At this point, Schneier is sure, firing 50% or 60% of the US federal workforce is just not possible: “That is not a thing. Today, it cannot be done.” Yes, he adds, there’s all this talk of AI agents that could in theory replace humans in government, “but we’re not there yet.”
Your email address will not be published. Required fields are markedmarked