Negotiating AI Contract Terms: A Practical Guide
A new guide outlines key contractual provisions for companies using or providing AI-powered services, covering liability, data usage, compliance, and insurance.
A guide for in-house and outside counsel identifies key issues when negotiating contracts involving artificial intelligence, whether as a customer or a vendor. It advises aligning any contractual commitments with the company's internal AI governance program and usage policies before negotiation. Key clauses discussed include restrictions on AI processing of confidential or personal data, requirements for enterprise-grade tools, and prohibitions on using counterparty data for model training. The guide also covers representations about compliance with specific AI-related laws, such as the EU AI Act or Colorado AI Act, which counsel should seek to narrow, rather than agreeing to broad "compliance with all laws" provisions.
As companies integrate AI into operations, contracts must evolve to allocate novel risks related to data security, algorithmic bias, and regulatory changes. Standard liability and indemnification clauses may be insufficient, creating unintended exposure. Counsel should update negotiation playbooks to address AI-specific risks, including limitations of liability, insurance coverage, audit rights, subcontractor oversight, and requirements for human-in-the-loop review, ensuring that business teams can operationalize the final terms.