Source: International homepage (paywall)
Anthropic's move to involve legal experts in AI agent training signals that autonomous deal-making requires more than language model capability—it demands system-level accountability for binding commitments. Enterprises want autonomous agents to close faster and cheaper, but legal frameworks assume human judgment and reversibility at key decision points. The gap is accountability: liability chains and negotiation guardrails must exist before autonomous agents make contractual errors that expose companies to litigation.