Tort Liability for Damages Caused by Artificial Intelligence Systems: Current Challenges and Future Prospects

Resumo

The contemporary era is marked by an unprecedented technological revolution in the field of artificial intelligence (AI), with autonomous systems progressively integrated into critical sectors including public administration, healthcare, transportation, and essential infrastructure in Algeria and globally. This massive integration raises profound legal challenges, particularly concerning the application of classical tort liability frameworks to autonomous and opaque algorithmic entities, notably deep learning systems that evolve beyond their initial programming. This article examines, through a comparative perspective, the adequacy of the Algerian legal framework (Articles 124, 125, 138, and 140 bis of the Algerian Civil Code), French law (Article 1242, paragraph 1, of the French Civil Code), and European Union regulations (the proposed EU AI Liability Directive) in addressing damages caused by AI systems. Employing a qualitative and comparative analytical methodology, the study demonstrates the structural insufficiency of the traditional elements of tort liability—fault (faute), causation (lien de causalité), and subjective accountability—when confronted with the decision-making autonomy and algorithmic opacity of the "black box." The article concludes that an imperative legislative reform is necessary to establish a strict liability regime grounded in risk theory, coupled with the implementation of a mandatory insurance system to guarantee full victim compensation and legal certainty in the context of digital transformation.