What to say to an owner who distrusts AI companies

What do you say to an owner who distrusts AI companies?

Agree with the instinct, then show the controls. Distrust of AI companies is reasonable, and arguing the owner out of it fails. What helps is being specific about what the owner keeps control of: the scope, the exclusions, the price, the use limits written into the agreement, and the final decision to sign or not.

This page is for partners who hear "I don't want to help AI companies" or "I don't trust big tech with our data" from an owner they know.

Separate the three kinds of distrust

Owners usually mean one of three different things. Ask which one, because each has a different answer.

Do not treat the third as a misunderstanding to fix. Some owners will say no, and that is a legitimate outcome.

The controls the owner keeps

• SourceX manages the transaction. The company deals with one party for sourcing, rights review, delivery and payment rather than negotiating with each buyer directly.
• Scope and exclusions are the company's. It decides which systems and periods are in, and which are out.
• De-identification and redaction are agreed first. The rules are set with the company before any work begins.
• Delivery follows signature. Data is delivered only after an executed agreement and the company's authorization.
• Ownership stays put. Data is licensed, not sold, typically exclusively for AI training for an agreed term.
• Nothing is binding until signed. The owner can stop at any point beforehand.

What the agreement is worth

A written promise is only as good as its enforcement. The FTC's staff guidance on confidentiality commitments says companies' promises not to use customer data for undisclosed purposes, such as training models, are enforceable, whether made in privacy policies, terms of service or other materials. That is staff guidance, not a rule, and it concerns promises to customers, but it shows regulators treat data-use commitments as binding. Owners should still have counsel read the contract. This is general information, not legal, tax or financial advice.

For a diligence list the owner can run on any buyer, see the due diligence checklist for vetting an AI data buyer.

What to say: three short scripts

If a headline triggered the worry, there is a script for it in how to answer an owner worried by AI data sale headlines. If the concern is about an unsolicited offer for individual files, see what to do when an AI data firm offers to buy client work files. For what happens to the records after delivery, point to what an AI buyer actually does with licensed records.

When to drop the topic

Stop when the owner gives a clear values answer, when they ask you not to raise it again, or when the conversation turns into persuasion. A partner's value is trust, and one push too many costs more than the introduction is worth. Rewards are not guaranteed; they are paid only after the buyer pays and SourceX receives its fee, so there is no reason to press. Timing questions are covered in whether to wait before licensing, and the full path from concept to deal is in how to sell data to AI companies. Sponsors can also read the reputational risk view for portfolio companies.

Next step

If the owner is open to a screen, register as a partner and make the introduction, or have the company apply at sourcex.si/apply. The referral earnings calculator shows how the partner formula works, and the FAQ covers program basics.