Meta updated its commercial terms to shift sole responsibility for AI feature risks to business users. Continued use after November 20, 2026 implies acceptance of this change. A proposed US AI Agent Accountability Act would extend liability for reckless AI agent damage. Forrester argues contracts must link governance intent to risk management for third-party AI. It advises bounding agent autonomy, demanding evidence over assurances, and matching accountability to control. Contracts should include real remedies like remediation, suspension, and financial recourse. Forrester will discuss these clauses at its Security & Risk Forum in Washington, DC on November 9–10.
What changed
Meta's commercial terms did not explicitly assign sole responsibility for AI feature risks to business users.
What it unlocks
Businesses can now demand contractual safeguards for third-party AI agent autonomy and accountability.
- November 20, 2026
Sources