Case Explained: AI and the Rise of Grey-Collar Crime: Who Bears Legal Responsibility? – Legal Perspective

With AI systems making autonomous decisions, existing laws struggle to define accountability—raising urgent questions about culpability, regulation, and rights in the machine age.
| Photo Credit: Alfieri/Getty Images

We have categorised crimes in every shade of human capability: white-collar schemes, blue-collar thefts, and every misstep in between. But what if the culprit isn’t human? Enter the next frontier: grey-collar crime, a murky middle ground born not of flesh and blood but of algorithms and silicon—a valley formed of rare grey metals and moral ambiguity. Welcome to the age of AI crime, where the question isn’t whodunit but whatdunit.

This new genre of grey-collar crime necessitates a brand-new legal framework. But where to start? Isaac Asimov’s (an American science fiction writer) three laws of robotics offered a framework to curtail the dangers of AI:

1) A robot may not injure a human being or, through inaction, allow a human being to come to harm;

2) A robot must obey the orders of human beings except where such orders would conflict with the First Law; and

3) A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

Also Read | Don’t panic about Ghibli-style AI art. Know what’s really at stake

But even Asimov acknowledged their limitations, later introducing a Zeroeth Law, “A robot may not harm humanity, or, by inaction, allow humanity to come to harm.” Yet, will such laws suffice for entities that evolve through self-learning, capabilities expanding beyond their creator’s intentions? AI’s self-learning capabilities complicate legal accountability: adapting through experience and straying from the original code, creators’ influence diminishes as autonomy grows. This mandates the existence of regulations and restrictions for grey-collar infractions.

With restrictions come consequences for breaking them. Consider a court case in the not-too-distant future: a self-driving car accident. It’s a stylised trolley problem, only you aren’t pulling the lever; a computer system is. It’s one thing if a person driving a car had to make a split-second decision between their life and another, but how can you code AI to make that decision? It raises a pertinent legal concern: who is culpable? The relatively new artificial intelligence (virtually a minor) or the parent company?

At what point does AI transition from property to progeny? Is eighteen years of existence universally the mark of adulthood? Rather than measuring AI’s autonomy in years, should a test, say the Turing Test, be used to create a statute of limitations? The Turing Test was created in 1949 to define when a computer’s ability to exhibit intelligent behavior was equivalent to that of a human. As even the simplest forms of artificial intelligence seem to stretch far beyond that of a human, does this decades-old test still serve as an adequate benchmark for AI’s autonomy? When code progresses so far that the parent relinquishes control and responsibility for their creations’ flaws, we reach the cliff of historical precedence. This is where grey-collar crime attempts to tackle what defies categorisation.

Star Trek tackles a similar conundrum as Data, the android, undergoes a trial to determine his autonomy in the episode The Measure of a Man. Through a three-part test, evaluating Data’s intelligence, self-awareness, and consciousness, his right to choose was upheld, challenging our definition of a machine. If AI successfully achieves a status similar to Data’s, our legal systems must evolve.

Is the rise of AI a modern-day Prometheus, mimicking the vicious spiral of Frankenstein and his monster, rushing society into rapid, uncharted advancements? A cautionary tale of the ethical boundaries of science pushed too far. Is Victor culpable for the hell unleashed by his creation? Likewise, who bears responsibility for AI’s transgressions? Will AI creators forever shoulder this burden, or will there come a moment when AI, like Frankenstein’s monster, outgrows its nascency?

Also Read | Can AI capture Miyazaki’s soul? OpenAI’s GPT-4o tests animation’s boundaries

A new issue arises of how and where to place blame. But moral blame is not the same as legal culpability—assigning legal culpability requires regulations on the creators of the software. Those restrictions can only be created and enforced when the AI must pass a strict moral, ethical, or social test, labelling it as a figure capable of assigning blame to. Just as the FDA stamps its approval on foods and drugs, an institution must exist to stamp its approval on artificial intelligence software, either labeling them as evolved enough to face consequences in court or elementary enough that their engineers are at fault. But who is creating these regulations? And who are the people regulating acts? No one, yet.

Grey-collar crime isn’t just a legal challenge; it’s a philosophical one. Are we ready to prosecute crimes committed that blur the line between tool and entity? The future requires answers as complex as the systems we’ve unleashed. In the words of Mary Shelley (English novelist), “You are my creator, but I am your master.” Now, it’s up to us to decide what justice looks like in the age of AI–and its various shades of grey.

Ezri Rohatgi is a graduating high school senior from San Diego, California. She is interested in studying international human rights law, with a focus on border conflicts and post-colonial exploitation.