Ethics in AI - Why it matters

Ethics is the study of the good, and of how best to act in the light of it. AI ethics is the study of how best to use, design and develop AI and of the many goods – welfare in all its many forms, justice, the pursuit of knowledge, autonomy, human connectedness – which bear upon those questions. AI ethics is thus concerned for example with ensuring that AI systems are transparent and free from bias, and with devising appropriate regulatory regimes for them by making sure liability is clear if things go wrong. And it is concerned with demarcating where AI can properly take over from humans and where – including in matters of ethical judgment – things are best left to us.

AI ethics matters because AI is very powerful. It can bring huge benefits, from medical diagnosis and scientific discovery to relieving people of repetitive work. But these powerful systems can also lead to harm, either unintentionally or in the hands of malicious agents. AI ethics helps to ensure that AI is developed in ethically sensitive ways and is deployed responsibly.

AI very likely will change human life profoundly, but that need not mean it will change it for the worse. There have been other convulsive technological changes in human history – writing, railways, the internet – and they are usually less terrifying in retrospect than they are in prospect, partly because those of us able to look back have been shaped by the new worlds these changes have brought into being. AI ethics can help us to think more clearly about such changes. Ultimately what the human future looks like is up to us. AI ethics helps us to shape that future in the light of what we value most.