Title: AI Is Automating Government—But Not Where It Matters Most
If you’ve ever fought with your medical insurance carrier, you already understand the frustration of dealing with automated systems. The same algorithms now examining your insurance claims are coming for your tax returns—and the consequences could be far worse.
The Problem with Automated Decision-Making
Whether it’s a hospital billing department or the IRS, organizations increasingly rely on artificial intelligence to process claims and returns. These systems operate on rigid logic: either 2 + 2 equals 4.0000, or your submission is rejected. There’s no room for nuance, context, or critical thinking.Here’s how it breaks down:
- Billing departments must code everything precisely, or insurers reject claims automatically.
- Doctors must navigate coverage rules that change constantly.
- You are left calling multiple departments, waiting on hold for hours, and hoping the person on the other end isn’t distracted by screaming children or household chores.
I recently spent hours contesting a bill. One representative deflected my questions for thirty minutes before I heard her child screaming in the background. After another hour on hold, the line went dead. Coincidence? I doubt it.
The IRS Is Next
The IRS plans to adopt this same model for examining tax returns. Every discrepancy flagged by an algorithm must theoretically be reviewed by a human—but will those reviewers be focused, or will they be working from home with dishes piling up and dinner to make?Meanwhile, the U.S. tax code contains approximately four million words—seven times longer than the Bible—and has grown 70 percent since the 1990s. Americans collectively spend 1.5 billion hours filing returns each year. The complexity is staggering, yet the government’s response is automation without accountability.
The Real Question
Here’s what frustrates me most: if companies like Palantir can build software to audit taxpayers, why aren’t those same tools being used to prevent the billions of dollars in fraud we only discover when a determined thirty-one-year-old investigates Learning centers in Minnesota?
A Better Path Forward
- End remote work for critical government roles. As Elon Musk bluntly called it, “pretend work” undermines accountability.
- Build taxpayer-facing AI tools. Let citizens plug numbers into an IRS website that references the tax code automatically—no paid preparers required.
- Deploy fraud-detection AI aggressively. Use the same technology auditing citizens to audit government spending and political corruption.
Americans shouldn’t need to pay hundreds of dollars and spend dozens of hours just to comply with a tax system designed to confuse them. And they certainly shouldn’t watch politicians enter office as bartenders and leave as multimillionaires while algorithms reject legitimate returns over decimal points.If we’re going to let AI run the bureaucracy, let’s aim it at the right targets.
