Anthropic is working with legal professionals to help train AI agents to negotiate and strike deals on behalf of humans. The effort focuses on how automated systems should operate within the practical and legal boundaries of negotiation, particularly when they act in real-world transactions.
Both outlets frame the move as a response to an emerging challenge: as AI agents increasingly take on tasks that involve bargaining, commitments, and interaction with counterparties, companies must address the negotiation “rules of the road.” The Financial Times emphasizes that negotiation is not only a technical problem but also one shaped by legal and regulatory constraints, which increasingly pull technology firms into collaboration with lawyers. India Today highlights the goal of teaching agents to negotiate more like humans, reflecting the need for better communication and decision-making during deal discussions.
Overall, coverage agrees that the initiative brings legal expertise into AI agent training to improve how agents manage negotiation and agreements. The sources differ mainly in emphasis—one on the legal framework and industry pressure, the other on the human-like negotiation capability being pursued.
