Anthropic is consulting with law firms to develop fair negotiation norms for AI agents that will haggle over prices, resolve complaints, and improve contract offers. The company initiated these consultations as it prepares for a future where autonomous systems actively cut deals on behalf of users. Experiments at the company have already demonstrated that stronger AI models are capable of negotiating better deals, raising concerns that this capability could inadvertently widen societal inequality.
To address these complex challenges, Anthropic is hiring legal professionals as Conceptual Reasoning Fellows to test its Claude models on judgment-heavy questions. These fellows spend four months evaluating the AI across complex domains including law, philosophy, economics, and political science. This initiative stems from a fundamental data hurdle in humanities and social sciences, where unlike mathematics and coding, verifiable training data is scarce and right answers lack clear-cut definitions.
Anthropic researchers note that defining quality in legal contexts is notoriously difficult, making expert human evaluation essential for building reliable evaluations. By bringing in legal expertise, the company aims to establish robust frameworks before autonomous agents become widespread participants in economic negotiations. This research ultimately matters because establishing ethical guardrails and fair norms now will prevent autonomous economic agents from destabilizing markets or deepening societal divides as they gain adoption.