As AI increasingly takes on decision-making roles once reserved for humans, a thorny question arises: can an algorithm ever truly be held accountable? And more provocatively—can it understand right from wrong?
By Stuart Kerr, Technology Correspondent
Published: 20 July 2025
Last Updated: 20 July 2025
Contact: liveaiwire@gmail.com | Twitter: @LiveAIWire
Author Bio: About Stuart Kerr
The Rise of Machine Morality
AI systems now evaluate parole candidates, screen job applicants, and even diagnose mental health conditions. As these decisions carry real-world consequences, society is being forced to confront a new dilemma: if a machine gets it wrong, who takes the blame?
Your article Black Box Medicine warned of this opacity. We’re building machines whose internal workings are often so complex, not even their creators can explain them. Yet we still let them make life-altering decisions.
The Stanford HAI initiative recently explored real-world cases where AI failed in healthcare settings—and no one was clearly responsible. If there is no human hand to slap, no soul to admonish, how do we assign justice?
Judgement Without Consciousness
We’ve built systems that can learn patterns, mimic ethical choices, and even apologise. But they don’t understand pain, joy, or consequence. That absence of consciousness is critical. Can an entity without awareness of suffering be said to make a moral decision?
Philosophy Eats AI argues that real morality requires more than good outcomes. It requires intention. And intention, so far, remains uniquely human.
In Digital Necromancy, you explored how algorithms reanimate likenesses of the dead. This raises eerie questions: Is it ethical to simulate someone’s identity if the system generating it lacks any concept of consent?
The Algorithm Will See You Now
AI isn’t just in the lab—it’s already judging us. In courtrooms, algorithms inform sentencing. In classrooms, they shape learning paths. In hiring, they filter who gets an interview. Your Algorithm Will See You Now piece documented how quietly this authority has spread.
The World Economic Forum lists moral agency as a top ethical issue for AI. Without clear regulations and explainable models, algorithmic decision-making risks becoming a form of moral outsourcing. We let machines judge—but they do so without values.
Towards a Moral Framework
So what’s the path forward?
Governments and institutions are beginning to draft ethical charters. The European Parliament has called for robust guidelines to ensure that autonomous systems are trained within moral boundaries. Their PDF, Moral Machines: Teaching AI Right from Wrong, suggests programming AI not only with safety protocols, but with ethical reflection mechanisms.
More foundational is the Stanford AI Bill of Rights, which proposes a framework for accountability, transparency, and redress. It stops short of suggesting AI itself can be punished—instead, it centres human responsibility.
The Illusion of Accountability
It may be comforting to imagine that AI will one day grow a conscience. But for now, every ethical decision it makes is merely a reflection of our own inputs, blind spots, and intentions. AI doesn’t experience guilt. It doesn’t hesitate before acting. It cannot repent.
Until that changes—if it ever does—we must resist the temptation to see algorithms as judges or guardians. They are tools. Sophisticated, powerful, even transformational—but still tools.
If we forget that, we don’t just risk injustice. We risk forgetting what it means to be moral ourselves.
About the Author
Stuart Kerr is the Technology Correspondent for LiveAIWire. He writes about artificial intelligence, ethics, and how technology is reshaping everyday life. Read more