AI provisions are appearing in clinical trial agreements, CRO master services agreements, vendor contracts, and data use agreements — often as hastily drafted boilerplate that neither party fully understands. This session provides practical negotiation guidance from both sponsor-side and site-side perspectives. We'll work through the clause categories that matter most: permitted and prohibited uses of AI in performing services, restrictions on inputting confidential or personal data into AI systems, training-data prohibitions, disclosure and transparency obligations, allocation of responsibility for AI-generated errors, intellectual property in AI-assisted work product, and audit rights. We'll review sample language, discuss common sticking points, and identify where a blanket prohibition creates more problems than it solves.

Takeaways

• Recognize the core AI clause categories now appearing in research agreements
• Evaluate sample language from both sponsor and site negotiating positions
• Anticipate downstream operational consequences of restrictive AI provisions