Download This PaperThe AI Accident Network: Artificial Intelligence Liability Meets Network Theory
Work published in 95 TUL. L. REV. (2020-2021).
58 Pages Posted: 22 Apr 2020
Date Written: March 26, 2020
Abstract
Network theory holds great unappreciated value when it comes to analyzing and confronting new legal issues induced by emerging technologies. It can assist regulators, judges and insurers to visualize multifaceted legal questions in a manner that assists them to better understand the complexity of new legal enquiries, and the appropriate policy intervention. A complex challenge the legal realm faces today is that of Artificial Intelligence (AI) liability. This Article will utilize network theory in this context. It will do so by combining network theory's features with Fletcher's nonreciprocal risks paradigm. This combination will provide essential tools which will show that a strict liability regime is the best-suited regime to apply when AI causes harm, and will provide indicators to identify the entity who should be held strictly liable.
Fletcher's nonreciprocal paradigm was heavily criticized upon its emergence. Calabresi's cheapest-cost avoider strict liability approach eventually became the prominent approach in the context of strict liability. However, this Article claims that today's network economy and networked society offers a new opportunity for Fletcher's paradigm to be applied given the new ways in which we interact and inflict damages upon each other, which were not feasible in the past. This Article integrates network theory into the field of AI tort law for the first time. It presents a new and unique methodology to the ongoing dispute about the appropriate liability regime that should apply when AI causes damages.
Keywords: Artificial Intelligence, Tort Law, Network Theory, AI Liability
Suggested Citation: Suggested Citation
Anat Lior (Contact Author)
Drexel University Thomas R. Kline School of Law ( email )
3320 Market Street
Philadelphia, PA 19104
United States
