“AI can help us make better moral decisions.”

Bernard Sinclair-Desgagné is emeritus Professor of Economics and Corporate Social Responsibility at SKEMA Business School. He talks to Glimpse about the implications of artificial intelligence (AI) in moral decision-making and the ethical challenges it raises. Interview.

By Kevin Erkeletyan

How do the ethical issues raised by AI differ from those raised by other technological advances?

The ethical issues surrounding AI are not fundamentally new. The same questions come up as when digitisation and mechanisation were introduced. However, AI presents particular challenges because it learns and evolves very quickly. Unlike a passive computer, AI learns on the job and can eventually create things, and this profoundly changes human-machine interactions.

You make a distinction between morality and ethics…

Yes. Ethics is theoretical, it’s a set of abstract, universal principles that often justify decisions after the event. Morality is the concrete application of this, and takes account of situational factors. By analysing these contextual elements, AI could be a major asset in making more informed decisions

What role could AI play in the governance of a “moral” company?

It has enormous potential. Company employees are often caught in moral dilemmas, having to choose between conflicting actions with significant ethical consequences. These dilemmas often lead to considerable emotional stress, which can hamper the ability to make rational decisions. AI can eliminate cognitive biases and incentives. Take whistleblowers, for example: AI could replace them or facilitate their action, because it is not influenced by career concerns or internal pressures. By analysing contextual elements impartially, it can support fairer, more ethical decisions.

You also mention a human tendency towards “moral accounting”…

Yes. If I make a good moral decision one day, I’ll tend to let myself diverge a little the next. It’s a very human thing to do: I ate vegetables the day before, so I allow myself an extra slice of cake the next day. AI can counterbalance this trend by offering a more objective, unchanging assessment of moral situations. Unlike us, it can maintain more consistent and rigorous ethical standards.

How can AI be integrated into these decision-making processes?

AI can be integrated into moral decision support (MDS) systems. These systems can include ethical alert algorithms, ethical recommendation systems and ethical performative algorithms. Ethical alert algorithms are designed to flag potentially problematic situations before they become crises. Ethical recommendation systems can offer advice on the best actions to take based on the data available, while ethical performative algorithms can automatically adjust actions according to pre-established ethical standards.

But isn’t it risky to rely on AI to make important decisions?

Yes, of course. There is no question of blindly following its recommendations. Despite its capabilities, AI is still prone to errors. So it’s crucial to maintain a critical approach and not lose our ability to think for ourselves. It is essential that people remain the f inal decision-makers, so that the decisions taken are in line with the company’s values.