AI Liability Along the Value Chain

In a new report for Mozilla, Beatriz Botera Arcila explores how liability law can help make sense of the complex nature of responsibility with respect to AI. "The multiplicity of actors involved in AI value chains represents a challenge for liability law because it is often difficult to know what each actor’s responsibilities should be and how they contribute to AI outputs. In accidents that involve an AI system, it is thus diffi- cult to identify the cause, and often the multiple mistakes or negligence, that led to the accident. This is a function of, and aggravated by, the opacity and technical complexity that characterizes AI systems."

Read the full report from Mozilla.