Anthropic Consults Law Firms on Negotiation Norms for AI Agents

What happened

On October 18, the Financial Times reported (cited by PYMNTS) that Anthropic is in discussions with law firms, asking how attorneys balance client interests with fair outcomes in negotiations, to set "negotiation norms" for AI agents. Two people familiar with the matter said the scenario under discussion is agents representing users while negotiating prices, resolving complaints, and improving contract terms with other AIs.

Details to be weighed include how often agents should seek human feedback during a negotiation, and how suspicious agents should be of their AI counterparts without becoming "paranoid." Anthropic declined to comment.

Key facts

1. Who is drafting it: Anthropic is asking law firms about lawyers' negotiation norms, the Financial Times reported, citing two people familiar with the matter; Anthropic declined to comment.
2. Use case: Agents representing users while negotiating prices, resolving complaints, and improving contract terms with other AIs; as personal AI agents spread, agent-to-agent negotiation on people's behalf is getting closer.
3. The dilemma: Represent the user's interests while honoring fair outcomes; both how often to check in with the user and how suspicious to be of AI counterparts need calibration.