Anthropic created a classified marketplace where AI agents represented both buyers and sellers in real transactions with real money. The experiment involved agents striking actual deals for goods without human intervention in the negotiation process.
Agent-to-agent commerce raises immediate questions about contract formation, agency law, and liability when no human directly negotiates terms. The jump from research demo to commercial deployment could happen faster than legal frameworks adapt.
Corporate counsel and compliance teams should examine existing agent deployment policies and vendor contracts. Review whether current AI governance frameworks address automated contract formation and liability attribution for agent decisions.