Could someone identify our company from a licensed dataset?
Yes, a licensed dataset can point back to its supplier if identifiers such as client names, product names, distinctive projects and email domains are left in. Redaction choices are agreed with the company before any work begins, and buyers generally do not need the supplier's identity inside the training copy.
Can a licensed dataset be traced back to our company?
It can be, unless identifiers are handled on purpose. Removing employee and customer names is only half the job: product names, client names, distinctive projects, internal tool names and email domains can all point back to the supplier. Which identifiers are removed or masked is agreed with the company before any work begins, and nothing is delivered without an executed agreement and the company's authorization.
Most anonymization advice is about people. This page is about the other question an owner asks: "will a buyer, or anyone who sees the data, know it came from us?"
What points back to a company
Run through the records the way an outsider would. The test is not "does this name a person" but "could a stranger work out whose business this is".
| Identifier | Where it shows up | Why it gives the supplier away |
|---|---|---|
| Email domains and signatures | Email archives, shared inboxes, support threads | The domain names the company outright |
| Client and vendor names | CRM notes, tickets, contracts, invoices | A short client list narrows the field fast |
| Product, feature and code names | Jira, pull requests, release notes, runbooks | Branded names are searchable |
| Distinctive projects or incidents | Post-mortems, project channels, board decks | A well-known outage or launch is a fingerprint |
| Locations and org structure | Office names, team channels, job titles | Unusual structures identify a firm |
| Internal URLs and ticket prefixes | Chat links, wiki pages, error logs | Subdomains and prefixes are unique |
Why buyers rarely need your name in the data
The value of a work record to a buyer is its structure: what was asked, which steps were taken, which tools were used, what the outcome was. A ticket thread teaches the same thing whether the product is called "Orion" or "Product A". In general, the training copy does not depend on the supplier being identifiable.
Two things are easy to confuse here. Identity inside the data is one thing; identity inside the contract is another. A buyer's reviewers may still need to know who the licensor is for rights and diligence purposes. What the owner can negotiate is who knows, under what confidentiality terms, and what the buyer may do with that knowledge.
The three entity-level choices owners weigh
For each category of identifier, the company usually picks one of three treatments. The list below describes the options to raise, not a fixed menu; the final rules are set in the agreement.
| Treatment | What it means | When owners pick it | Trade-off |
|---|---|---|---|
| Remove | The name or passage is deleted | Highly sensitive clients, regulated counterparties | Threads can lose context |
| Replace consistently | "Client A" stays "Client A" across the whole set | Names that matter for workflow continuity | Needs a careful mapping that the buyer never sees |
| Keep | The name stays | Public facts such as a widely published product name the owner is content to be linked with | The record is easier to attribute |
Consistent replacement is often the best balance, because multi-step workflows stay readable while the real names disappear.
Questions to settle before anything is signed
Use this as the owner's checklist when the redaction rules are drafted.
- Which identifier categories (company, clients, vendors, products, people, locations) are removed, replaced or kept?
- Who checks the redaction before delivery, and can the company sample the output first?
- Does the agreement bar the buyer from trying to identify the supplier or any individual?
- Who at the buyer knows the supplier's identity, and under what confidentiality terms?
- Are any clients' contracts restricting disclosure of their names, and do those records need to be excluded?
- Are the most distinctive systems or time periods better left out of scope?
What to say to an owner who worries about this
For the broader set of owner worries, point them to common concerns about licensing company data and the explanation of how company data is anonymized before AI licensing.
Limits: no process makes re-identification impossible
De-identification lowers the risk of attribution; it does not remove it. The more distinctive a company, with one famous client, one signature project or a very small set of customers, the more it needs masking or exclusion. Linking a licensed set with other data a reader already holds can also narrow the field.
If the owner's competitive position depends on a specific project staying secret, exclude it from scope or do not proceed. Nothing is binding until the company agrees price and terms and signs. The legal side of an owner's exposure is covered in a realistic risk map for licensing records, and the buyer's side in what an AI buyer does with licensed records. Owners who read licensing as a distress signal can start with whether licensing data means a company is failing.
This is general information, not legal, tax or financial advice. Confirm with your own counsel before acting.
Next step
If an owner you know is hesitating mainly on identifiability, give them the checklist above and offer to introduce them. You never see or describe their records; you only make the introduction. Register as a partner, or see how the reward formula works in the referral earnings calculator and the FAQ. The company can also apply directly at sourcex.si/apply.
- Step 1Share your linkSend your personal link to a company you know.
- Step 2Company appliesThe company applies itself at /apply.
- Step 3Buyer selects and paysThe buyer selects and pays for the data and SourceX receives its fee.
- Step 4You get your rewardYour share of SourceX fees becomes payable.
Common questions
Is removing employee names enough to keep the company anonymous?
No. Employee names are only one identifier. Email domains, client names, product and code names, internal URLs and well-known projects can each reveal the supplier. The company decides, before any work begins, which categories are removed, replaced with consistent placeholders or kept.
Will the buyer know our company's name?
That depends on the agreement. Identity inside the data is separate from identity inside the contract. Buyer reviewers may need to know the licensor for rights and diligence, so owners should negotiate who is told, under what confidentiality terms, and what the buyer may do with that knowledge.
Can we leave out our most distinctive systems or projects?
Yes. The company sets scope and exclusions, so a signature project, a sensitive client's records or an entire system can be left out. Excluding the most distinctive material is often the simplest way to lower the chance that a dataset is attributed to the company.
Can anonymized data ever be re-identified?
It can lower the risk but not eliminate it. Linking a licensed set with other information a reader already holds can narrow the field, especially for companies with few clients or a famous project. That is why the agreement should restrict attempts to identify the supplier or individuals.
Do I as a partner see any of the company's records?
No. Partners make introductions and give basic fit information only. They never export, upload or describe confidential records, and data is delivered only after an executed agreement and the company's authorization.
Related pages
- SourceX referral program frequently asked questions
- Common concerns about licensing company data, answered
- How is company data anonymized before AI licensing?
- Could licensing our records get us sued? A realistic risk map
- Does licensing data signal a company is failing?
- What does an AI buyer actually do with licensed company records?
Free resources
- AI readiness assessment — Ten questions, five dimensions, a score out of 100.
- EBITDA calculator — Reported and adjusted EBITDA from net income.
- MOIC calculator — Multiple on invested capital from realized and unrealized value.
- All free tools · MCP resource center
By SourceX Partnerships Team · Published 2026-10-09 · Updated 2026-10-09
Know a US company with valuable proprietary data?
Become a referral partner from anywhere we support, get your link and introduce an owner or authorized decision-maker.
Refer a company →I own a business
Explore licensing your company's data to AI developers worldwide. Start a short assessment; no uploads needed.
Start an assessment