Agentic AI systems are being deployed at scale in enterprise environments, yet the legal framework for determining who bears responsibility when these systems cause harm remains fundamentally unresolved. This article argues that agentic AI produces a categorically distinct form of harm called trajectory-based harm, which existing negligence, product liability, and vicarious liability doctrines cannot address because they assume that harm is traceable to a discrete human decision rather than an autonomous action sequence. Drawing on empirical data indicating that 80.9% of surveyed technical teams have moved past planning into active testing or production of AI agents, while only 14.4% report full IT and security approval for all agents going live, and that 74% of surveyed organizations expect to use AI agents at least moderately by 2027 despite only 21% having mature governance frameworks for autonomous agents, this article demonstrates that businesses face material liability exposure from agentic AI deployment that existing legal frameworks cannot allocate. The article analyses cross-jurisdictional regulatory responses, including the European Union AI Act and the Singapore Model AI Governance Framework, and proposes a practical compliance architecture for businesses that includes structured deployment screening, constitutional checkpoints, mandatory trajectory logging, and contractual liability allocation. The analysis has direct implications for corporate counsel advising on agentic AI procurement, deployment, and risk management across multiple jurisdictions.
Business Law Review