Anthropic has consulted law firms about AI agents that negotiate for users, the Financial Times reported Sunday. The talks cover haggling over prices, customer-service disputes and even salary offers. One idea is signed user documents that let an agent check another agent's orders without seeing its full prompt.