Tesla Robots, Google AI and Criminal Law: Who Is Responsible When a Robot Commits a Crime?

Technology companies such as Tesla and Google continue to push artificial intelligence and automation into areas that once seemed like science fiction. But as robots and AI become increasingly capable, Attorney Michael Kotik raises an interesting criminal-law question: What happens if you tell a robot to commit a crime and the robot actually does it?

Imagine that you own a Tesla robot and intentionally instruct it to enter a store and steal merchandise. Can the Tesla robot be arrested? Can you be arrested?

A robot is a machine, not a human criminal defendant who can ordinarily be arrested, prosecuted, and sentenced to jail. That does not mean, however, that the person directing the robot would escape potential criminal liability. If a person intentionally uses a robot as the instrument to commit a theft, prosecutors could potentially focus on the person who gave the instructions, exercised control over the machine, and possessed the required criminal intent.

What Does Pennsylvania Law Say About The Analysis?

In Pennsylvania, the analysis would depend on the particular criminal statute and the specific facts. The fact that a person’s own hands never physically touched the stolen merchandise would not necessarily prevent criminal liability. Courts routinely examine intent, participation, causation, and the actions of everyone involved when determining criminal responsibility.

Artificial intelligence creates an even more complicated question. Google and other technology companies have demonstrated how rapidly AI systems are advancing. What happens when a future robot does something that its owner did not specifically instruct it to do? If an autonomous system interprets an instruction incorrectly or independently takes an unlawful action, determining who possessed the necessary criminal intent could become significantly more complicated.

DUI Laws Around Automated Vehicles

There is also another interesting hypothetical: Could someone get a DUI while controlling a robot while intoxicated?

Pennsylvania’s DUI law applies to driving, operating, or being in actual physical control of the movement of a vehicle. Simply telling a household robot to take out the trash while intoxicated would not automatically constitute DUI. But as robots become larger, mobile, capable of transporting people or property, and increasingly similar to traditional vehicles, courts and lawmakers may eventually have to address where the legal boundary exists.

According to Attorney Michael Kotik, the larger criminal-law issue is ultimately one of control and intent. Technology does not necessarily create a shield from criminal responsibility. If a person intentionally directs a machine to commit a crime, prosecutors may look beyond the technology and examine the conduct and intent of the human being controlling it.

The truly difficult legal questions may arise when artificial intelligence becomes sufficiently autonomous that the machine’s actions can no longer easily be traced to a specific human command. As companies such as Tesla and Google continue developing increasingly sophisticated technology, legislatures and courts will likely face new questions involving intent, causation, control, and criminal responsibility.

Until then, one thing should be clear: telling a robot to commit a crime should not be viewed as a loophole that automatically protects the person behind the machine.

Disclaimer: This article is for educational and hypothetical discussion only and does not constitute legal advice. Tesla and Google are referenced solely for purposes of discussing developments in technology and artificial intelligence. Neither company is affiliated with or has endorsed Attorney Michael Kotik or this article.