Can AI Testify Against You? From Police Reports to Robot Officers

This Isn't Science Fiction. It's Today's News.

Two headlines from the past few months should concern every American who values their Constitutional rights:

At the same time, the Electronic Frontier Foundation (EFF) released an investigation into Axon's Draft One—an AI system that writes police reports from body camera audio. Their finding? The system is deliberately designed to prevent anyone from knowing what the AI wrote versus what the officer wrote.

Defense attorneys can't challenge what they can't see. And if you can't challenge the evidence against you, what happened to your Constitutional rights?

The Present: AI Already Writing Police Reports

Axon's Draft One launched in April 2024 and has become the company's fastest-growing product. It works by processing body camera audio and generating a written police report narrative in seconds. Over 100 police departments now use it.

Here's the problem: Police reports aren't just paperwork. They're used at bond hearings—often without live testimony—to determine whether you sit in jail or go home. A single word like "aggressive," "non-compliant," or "furtive" can tip the scales against you.

When officers write their own reports, the physical act of writing encodes the events into their memory. They can later testify from personal recollection. But when AI writes the report, that mental exercise is eliminated. The officer's memory fades faster, yet the AI-generated words remain—treated as gospel truth.

Some jurisdictions are pushing back. King County, Washington (which includes Seattle) has instructed police not to use AI for report writing. Utah is considering legislation requiring disclosure when AI generates police reports. But in most of the country, including Florida, there's no requirement to tell you that a machine wrote the narrative being used against you.

The Hidden Bias Problem

Axon claims that Draft One will "remove bias" from police reports by using neutral language. This sounds good—until you think about what it actually means for defendants.

This makes bias more insidious, not less. The prejudice that may have influenced the arrest still exists—you just can't see it anymore.

The Trajectory: Where This Is Heading

AI writing police reports is just the beginning. Consider the trajectory:

The 6th Amendment Collision

The Sixth Amendment to the United States Constitution guarantees:

This right exists because cross-examination is the "greatest legal engine ever invented for the discovery of truth." When a human witness testifies against you, your attorney can probe their memory, expose their biases, challenge their perceptions, and reveal inconsistencies.

But what happens when the "witness" is an AI system or a robot?

The Questions Nobody Has Answered

The Courts Are Already Struggling

In June 2024, the U.S. Supreme Court decided Smith v. Arizona, holding that the Confrontation Clause prohibits a "surrogate" analyst from presenting another analyst's forensic findings—even as the basis for their own expert opinion. The absent analyst must be available for cross-examination.

But what about machines? The Vanderbilt Journal of Entertainment & Technology Law examined this question in "The Future of the Confrontation Clause: Semiautonomous and Autonomous Machine Witnesses." Their finding:

The Harvard Law Review echoed this concern in October 2024 with an article titled "Machines Need Not Testify." The Confrontation Clause was designed for human witnesses. As machines become more autonomous, the gap between Constitutional protection and technological reality grows wider.

Florida's Enhanced Privacy Protections

Florida's Constitution provides privacy protections that go beyond the federal Constitution. Article I, Section 23 establishes an explicit right to privacy:

Florida courts have historically interpreted this provision to provide greater protections than federal law requires. As AI-generated evidence becomes more common, Florida defendants may have unique arguments based on state Constitutional protections that don't exist elsewhere.

The question remains open: Does an AI-generated "observation"—created by analyzing body camera footage and applying algorithmic interpretation—meet the legal standards for evidence? Can it establish probable cause when no human actually perceived the events it describes?

The Accountability Gap

The fundamental problem with AI in criminal justice is accountability. Consider what AI cannot do:

Frequently Asked Questions

The Question We Must Answer Now

We're not asking "if" AI will transform criminal justice—we're deciding "how." The rules we make today will determine the courtroom of 2035. Will defendants retain the right to confront their accusers, or will algorithms become unquestionable witnesses?

The Confrontation Clause was written when the Founders could not have imagined machines that generate evidence, robots that make arrests, or algorithms that determine freedom. But the principle behind it—that the accused has the right to challenge the evidence against them—is timeless.

As AI becomes more prevalent in law enforcement, that principle faces its greatest test. The question isn't whether technology will advance—it's whether our Constitutional protections will advance with it.