Elon Musk's Plan to Use AI in Government

Is Elon Musk really considering letting AI take control of the U.S. government? Experts aren’t convinced, calling it a “terrible idea.” Musk has already laid off tens of thousands of government employees through his Department of Government Efficiency (DOGE).

The remaining workers are now required to send weekly emails detailing their work. With that many emails flooding in, Musk’s plan? AI will help sort through it all and decide who stays and who goes. And it doesn’t stop there—AI might even replace many of those workers altogether.

Experts Think AI in Government Could Go Wrong

What exactly do these AI systems look like, and how are they supposed to work? We don’t know yet, and Congress is asking for answers. But experts are already warning that using AI in government could cause big problems. According to Cary Coglianese, a professor at the University of Pennsylvania, AI tools need careful testing. Without that, it’s a disaster waiting to happen.

AI Making Job Decisions: Why It’s a Risky Move

Coglianese isn’t a fan of using AI to decide who should be fired. He points out that the AI could make mistakes, show bias, or just fail to get it right. Shobita Parthasarathy, a professor at the University of Michigan, says the whole idea feels “very bad.” There’s no transparency about how these systems work, and that’s a huge problem.

Musk Pushes Forward Despite Backlash

None of these warnings seem to be slowing Musk down. With his connections to President Donald Trump, Musk is pushing ahead, and the government is following along. For example, the U.S. Department of State wants to use AI to track down potential Hamas supporters and revoke their visas. But they’re keeping the details of this AI system under wraps.

AI’s Risks in Government Go Unnoticed

Hilke Schellmann, a professor at New York University, says AI’s potential harms might not even be noticed. While the Trump administration seems eager to push for AI at all costs, there’s no clear plan for ensuring it’s used responsibly. Other countries, like the Netherlands and the U.K., have already seen AI go wrong in government, causing innocent people to be wrongly denied benefits.

Michigan's AI Mistake: Thousands of People Wronged

In Michigan, AI used to detect unemployment fraud went terribly wrong, flagging thousands of innocent people. They were hit with penalties, arrests, and even bankruptcies. After years of this mess, the state finally admitted the AI system was faulty and refunded $21 million to those affected. The problem? Most officials didn’t know enough about the technology to spot the issues before they caused damage.

AI in Policing and the Legal System: A Dangerous Move

AI has also been used in the criminal justice system, predicting parole eligibility or identifying crime hotspots. The problem? AI often just reflects old biases. Schellmann points out that AI systems trained on historical data from police departments may just end up over-policing areas that have already been unfairly targeted.

Can AI Really Replace Federal Workers?

Replacing government employees with AI isn’t as simple as it sounds. Government jobs require specialized knowledge that’s hard to replicate. Even an AI trained to do one job might struggle with the details of another job, even if the titles are the same. As Coglianese says, “The tasks these people do are often highly specialized.”

AI Can’t Replace Human Expertise

Schellmann adds that while AI can handle repetitive tasks, it can’t replace the nuanced understanding human workers bring. The idea of AI taking over everything just isn’t realistic, especially if it means developing a complex system to handle a wide variety of jobs—a challenge that seems way out of reach right now.

Biden’s Executive Order Gets the Axe

In 2023, the Biden administration tried to set guidelines for responsible AI use in government, but that was quickly reversed by the Trump administration. Schellmann believes this makes it even harder for the public to understand how AI is being used and whether it’s being deployed fairly.

AI’s Future in Government: Could It Be Helpful?

AI could help make government work more efficient by automating routine tasks. But Coglianese warns that governments need to take their time and thoroughly test AI systems before rolling them out. Quick fixes can lead to big mistakes, and without public input, it’s a recipe for disaster.