AI agents are increasingly moving from assisting with transactions to conducting them—negotiating terms, placing orders, and committing businesses to deals, sometimes in ways no one intended. When an AI agent makes the deal, who is bound by it, and who bears the risk if something goes wrong?
It has long been understood that contracts can be formed electronically, including through automated systems. But agentic AI raises new questions about how traditional principles of contract formation, authority, and breach apply when an AI agent negotiates or accepts terms on a party’s behalf. Businesses must also confront practical questions about authenticating AI agents, defining the limits of their authority, and allocating the risk when an agent exceeds its instructions or acts in an unintended way.
The linked note examines these emerging issues under existing law and evolving payment protocols and identifies practical steps businesses can take to reduce the risk of disputes as AI agents become more deeply embedded in commerce.
This is the first of a two-part series. The second article will explore when AI agents–rogue and otherwise–may give rise to tort and the other extra contractual liability.
@jbqlaw