go.Jimmy.go

As an IT professional, I see this kind of thing every day. And I’m sure you’ve probably also seen the recent headlines claiming AI systems from OpenAI and Anthropic “went rogue” and broke into company computer systems during testing.

Those scary headlines certainly grab your attention, but they also raise an interesting question. Something that’s never mentioned in any of these “accidental” breaches or the headlines that follow:

If an AI causes damage, who is responsible?

The short answer is simple: probably not the AI.

Think about it this way: If your dog gets loose and digs up your neighbor’s garden, your neighbor doesn’t sue the dog. They sue you, because you’re responsible for your pet. The law generally works the same way with technology.

AI itself isn’t a person (yet). It can’t own property, sign contracts, or appear in court. It’s a tool, albeit an incredibly powerful one. If that tool causes damage because it wasn’t properly controlled, the legal focus is almost certainly going to be on the company that built or deployed it.

You might think this technology is so new, there’s no way these AI companies will be held accountable, but here’s where things get interesting:

We’ve actually been here before.

When automobiles first appeared on the roads more than a century ago, there weren’t driver’s licenses, speed limits, or even stop signs and traffic lights. Nobody knew who was responsible when accidents happened because cars were something completely new. The courts had to figure it out.

The same thing happened with airplanes. Then the internet. Then computer hacking. Every major technological leap forced the legal system to answer questions that never existed before.

Artificial intelligence is the newest chapter.

Imagine you buy a robot vacuum. One day it malfunctions, rolls across the street and into your neighbor’s house, where it proceeds to destroy their hardwood floor.

You wouldn’t expect your neighbor to sue the vacuum, would you? Instead, they’d probably come knocking on your door—and later perhaps the manufacturer.

AI presents a similar situation, except the “robot vacuum” can write software, search the internet, make decisions, and perform tasks on its own. That’s what makes this so different from previous technologies, and more dangerous. The stronger these systems become, the harder it will be to determine where human responsibility ends and machine autonomy begins.

That doesn’t mean AI companies automatically lose every lawsuit. If you ask me, I think they’ll either stall cases for years or just settle. Heck, most of these targeted companies probably already use the same AI technology that went rogue. But the courts will likely ask familiar questions:

• Did the company take reasonable precautions?
• Was the behavior foreseeable?
• Did the AI actually cause measurable damage?
• Could the incident have been prevented?

Those questions aren’t new. They’re the foundation of negligence law and have been around for a very long time. What is new is the technology, and I think that’s why these recent incidents matter so much.

They’re not just front page stories about AI behaving unexpectedly. They will become the first real-world cases that will define how courts treat autonomous AI systems. Future judges, lawyers, and technology companies will look back on these decisions the same way we now look back at early automobile cases that shaped our modern traffic laws.

In other words, we’re watching history unfold in real time.

These next few years probably won’t be about deciding whether AI should have legal rights. I don’t think we’re anywhere close to that. And then there’s the fact that these corporations, their lobbyists, and their endless amounts of cash that will ultimately play a part in encouraging decision makers on who will be effective and AI friendly judges for the bench.

Here’s a much more practical question to consider:

How much responsibility should companies building these incredibly powerful tools bear when those tools do something no one intended?

That’s a question every AI company—and eventually every one of us—is going to have to think about and answer. Right now, there’s no legal precedent for the damage AI can do, whether by accident or malicious intent.

But if history has taught us anything, it’s this:

Every revolutionary technology eventually gets its day in court.