Our clients are starting to ask if we use AI to handle their accounts, and some are demanding we sign contract addendums banning it. How do we write a client-facing AI policy that satisfies their security concerns while keeping our team's AI-driven efficiency gains?

Question

Tyler Smith · Accepted Answer

When clients demand you sign contract addendums banning AI, do not panic and do not sign away your operational leverage. Most clients are not actually afraid of AI; they are afraid of their proprietary data being leaked into public training models or their intellectual property being compromised. To solve this, you need a clear, client-facing AI safety policy that aligns with your internal team guidelines.

Your policy must state exactly how you use AI. Specifically, declare that your team is prohibited from inputting any client-identifiable data, trade secrets, or proprietary code into public, consumer-grade AI models. Clarify that any tool used by your team is secured via enterprise-grade APIs where data is explicitly excluded from model training. This distinction is critical because API-based data usage keeps client information completely private and secure.

Next, update your standard master service agreements to reflect these protections. Rather than banning AI, your contract should guarantee that all data processed through third-party automated tools remains subject to the same strict confidentiality and security protocols as your traditional operations. This proactive approach turns a potential legal roadblock into a sales advantage. It proves you run a modern, highly secure, and efficient business. Introduce this as an Issue in your next Level 10 Meeting to get your leadership team and legal counsel aligned on the exact wording.