Updated October 2026
Can a competitor get my data?
You worry a rival could end up with your quotes, your prices or your client list. Not directly, if your license says so, and some buyers promise never to sell to your competitors. The real risks are a buyer reselling a dataset and a model learning how you price, and you close both with exclusions and use limits in writing.1,2
How it could happen
Three ways your records could reach a rival
Resale of the dataset. A broker or buyer that licenses your records may sublicense them, and some standard terms allow exactly that. If the license allows it and does not limit who, a dataset could end up with a company in your trade. MIT Technology Review found sellers often cannot even see which AI lab ends up with their data.3,4
The model itself. The data trains or tests AI models that anyone, including your competitors, may use. A model that has read thousands of your quotes may get better at quoting in your trade. It should not repeat your documents, but it learns how the work is done. That is the point of the deal: buyers want the workflows your employees repeat.5
Something left in. A price sheet or a client list that slips past cleaning. Buyers say client names are swapped for codes and personal details removed, but names of suppliers, prices and margins are business details, not personal ones, and may be left in unless you ask.6,7
What buyers promise
The promises in the ads
The promises vary: never sold to your competitors, sent only to AI developers named in your agreement, your company never named without written consent, or company secrets kept safe. Some buyers make none of them in public. Ask how each promise is enforced and whether it appears in the agreement you sign.1,8,3,2,9
What to do
Four ways to close the door
Leave out the crown jewels: your pricing workbook, bid tabulations, margin reports and anything a rival would pay to see. Buyers say any folder can be excluded, and sensitive categories are often left out already. Limit sublicensing to named AI labs, with no resale as a dataset. Ask for a no-competitor clause with a definition of competitor. Review cleaned samples before anything is released.10,11,12
Weigh it plainly. If your edge is your relationships and your crew, licensing old emails and tickets gives little away. If your edge is a pricing method nobody else has figured out, keep that part home. The quiz introduces you to the buyers we work with; put these four asks to each of them before you sign.
FAQ
Questions, answered with sources
Not as your files, but the model learns how the work is done, which is what buyers pay for. Keep out anything that is your edge, such as a pricing method.5
Not always. MIT Technology Review reported that sellers often cannot see the end buyer. Ask for the labs to be named.4
Every business is different, so this one is best talked through with the buyer you're working with. Their team goes over it with you before anything is signed.
Before you sign
Questions to talk through with your buyer
Every business is different, so these are best talked through with the buyer you're working with. Their team goes over them with you before anything is signed.
Questions for this kind of business
- How is "competitor" defined, and what happens if the buyer licenses our data to one?
- Can the buyer sublicense or resell the dataset, and to whom?
- Do customer, vendor or platform contracts, NDAs or API terms restrict this?
- Could the data include privileged material or trade secrets?
Questions for any buyer
- Is the license exclusive, and for how long?
- Can you sublicense the data, and to whom?
- Which AI labs will receive it, and will you tell us?
- How long do you keep it, and how is it deleted?
- What is your liability cap, and who indemnifies whom?
- Will the contract ban re-identification of people in the data?
- Who removes personal data, by what method, and do we approve it first?