Anthropic explores AI negotiation rules with law firms

Anthropic is collaborating with law firms to establish guidelines for AI agents to negotiate on behalf of users, aiming to secure the best possible outcomes. The company is seeking legal advice on how to ensure these AI systems act in the user's best interest while maintaining fairness. A source close to the project noted that Anthropic is developing 'negotiation norms' for AI interactions, particularly as these systems increasingly represent users in scenarios involving conflicting entities. The firm is also considering the extent of human oversight required during negotiations and how to balance vigilance against other AI systems without fostering distrust. In parallel, Anthropic reported a significant shift in its workforce composition, with a decline in junior technical roles since early 2025. The proportion of junior technical staff dropped from 21% to 14%, and new hires in this category fell from 27% to 9%. Additionally, the company is exploring a proposal that would require users to sign documents confirming their instructions and the rules governing AI agents.