On October 18th, according to the Financial Times, Anthropic is engaging with legal experts to explore a new issue: how AI agents can secure the best deals for users through negotiation. With the development of personal AI agents, AI companies envision a future where different agents can communicate with each other and negotiate on behalf of users. Two sources familiar with the matter revealed that Anthropic is seeking advice from law firms to understand how lawyers can protect clients' interests during negotiations and ensure fair outcomes. The sources stated that the company has been studying the 'negotiation norms' that AI models should follow when interacting. As agents become more capable, they will encounter other agents with 'conflicting goals' when acting 'on behalf of users.' Anthropic is also considering the extent to which agents should solicit human feedback during negotiations and the level of caution they should exercise when dealing with other AI agents to avoid becoming 'paranoid' or falling into 'conspiracy theory-like thinking.'