What happens when machine learning — computer programs that absorb^(a) new information and then change how they make decisions — causes investment losses, a car accident, or a wrong cancer diagnosis?

The big difference between machine learning and previous^(b) digital technologies is the ability to independently make progressively complex decisions — such as which financial products to trade — and continuously adapt in response to new data. But these algorithms don’t always make ethical or precise^(c) choices.

The imperfections of machine learning raise another important challenge: risks deriving from things that aren’t under the control of a specific business or user. Ordinarily, it’s possible to draw on reliable evidence to reconstruct the circumstances that led to an accident. But because machine learning is typically inserted within a complex system, it will often be unclear what led to a breakdown — which party, or “agent”^(d) (for example, the algorithm developer, the system deployer, or a partner), was responsible for an error and whether there was a problem with the algorithm.

Actually,^(e) accidents or illicit decisions can occur even without negligence on anyone’s part — as there is simply always the possibility of an inaccurate decision.

(https://hbr.org. Adaptado)