.jpg)
Does your vendor agreement allow customer data to be used for AI model training? Could a blanket AI prohibition block features your business actually wants? As vendors embed AI and machine learning into their products, in-house counsel need contract language that distinguishes between using AI to deliver a service and using customer data to train or improve models.
In this recorded session, Jonathan Perkel, Chief Legal Officer at Esusu, joined host Yishai Dvash, Head of Marketing at Chamelio, to examine key provisions governing vendor AI use. They analyzed data usage and model training clauses, aggregated and de-identified data carve-outs, input and output ownership, third-party AI providers in subprocessor chains, and confidentiality gaps. The session walked through a practical negotiation playbook with model language and redline examples, focusing on how to protect company data while preserving intended functionality.
They also discussed how legal teams can work with procurement, security, and privacy stakeholders to identify AI features and data uses early—before the contract reaches counsel’s desk.
Key Takeaways
- Distinguished service delivery from model training: Examined where contract language permits—or restricts—different uses of customer data.
- Spotted gaps across related provisions: Evaluated how data carve-outs, ownership terms, confidentiality obligations, and third-party AI providers affect contractual protections.
- Negotiated targeted restrictions: Explored model language and redline approaches that address unwanted data use while allowing the AI features your business needs.
- Surfaced issues earlier: Addressed workflows for coordinating with procurement, security, and privacy teams to clarify vendor AI practices before contract review begins.
Who This Session Was For
In-house counsel responsible for commercial agreements, technology procurement, privacy, and AI governance, along with legal operations professionals involved in vendor intake and contract review.
This program was offered free of charge courtesy of sponsor Chamelio. The recording is now available for on-demand viewing via our paid CLE library, In-House Connect On-Demand.
