Why business owners are hearing about AI data supply, and how to tell curiosity from a willingness to license: three dated signals, a six-rung ladder and a repeat-interview panel

Short answer

Owners are being drawn into the AI supply conversation from three directions that can be checked: AI developers publicly asking organizations for data that is not openly available online, regulators warning companies not to quietly rewrite their terms to use people's data for AI training, and data buyers announcing budgets for operational records. Whether owners are now more willing to license has not been measured. SourceX has not published owner interviews, so this page offers a ladder that separates curiosity from intent and a panel design to test the trend.

Visual briefFull-size diagram
A six-step staircase from raising the topic to signing, with the first three steps shaded as curiosity and exploration and a marked line at step four, the CEO's yes, where willingness begins; three panel waves are marked beneath.
The rungs and waves follow the method on this page; no interview results are shown because none are published.

For a business network owner, the useful question is narrower than "are owners interested in AI?" It is whether the member who just asked you about it is curious, worried or ready to talk to someone who would license their records. Those three people need different answers, and only the third needs an introduction.

Three dated signals that put the question in front of owners

Each signal is real and sourced. None of them is evidence about owners' intentions.

SignalSource and dateWhat it showsWhat it does not show
AI developers asking for dataOpenAI's Data Partnerships page, an older page whose terms may have changedOpenAI says it wants large datasets that reflect human society and are not already openly available online, and excludes sensitive personal data and third-party information (paraphrase)That any company's records will be bought, or any connection with SourceX
A regulator on quiet terms changesFTC staff post on Tech@FTC, February 2024Adopting more permissive data practices, such as using consumers' data for AI training, and disclosing it only through a surreptitious, retroactive change to terms may be unfair or deceptive (paraphrase)A rule or a decided case on this point; it is a staff blog post
A buyer with a budgetmicro1, announced 2026-10-09A stated 12-month commitment of $1 billion to acquiring and licensing enterprise operational dataThat the money has been spent, or that any given company's data will be bought

Each signal reaches an owner by a different road, so each tends to start a different first question. This is our reading of the signals, not interview evidence. An AI developer's call for data raises "are records like ours what they mean?" A software vendor's terms update raises "is someone already training on our data, and did we agree to that?" A budget headline raises "is there money in our archive?" The second question is about protection and the third about price. Only the first is about fit, and fit is the only one that leads to an introduction.

A six-rung ladder from curiosity to a signed license

The ladder turns a vague sense of interest into behavior you can observe and date. Willingness starts at rung 4, because under the program the CEO authorizes a licensing discussion.

RungWhat you can observeWhat it counts asWhat does not count
1. Raises the topicThe owner asks about AI and company records, unpromptedCuriosityAttending an event or liking a post
2. Asks about their own recordsNames a record type or system: "would our dispatch history count?"Curiosity with a targetA general question about AI
3. Shares fit factsConfirms full-time headcount at peak, years of records, systems and who owns them, at metadata levelExplorationOffering files, samples or screenshots, which a partner declines
4. The CEO says yesThe CEO, or an owner who is the CEO, agrees to a conversation with SourceXWillingness beginsA senior manager's enthusiasm on the CEO's behalf
5. The company appliesThe company submits its own applicationCommitment to a reviewA partner filling in a form for them
6. The company signsA Master and a Dataset Deal Confirmation that fixes the dataset and payoutA licenseAn indicative range, which stays indicative until the Confirmation

Two readings follow. A jump from rung 1 to rung 4 in the days after a headline, followed by silence, looks like a response to the headline. A slower climb that holds looks like changed interest. And the partner's part ends early: after rung 4 the company applies and decides for itself, and the partner's job is the dated record of the introduction.

Why some owners stay unconvinced

Hesitation is often well founded, and the program's own criteria explain much of it.

  • The best material is the clients'. SourceX's screening page lists material that belongs to the company's clients as usually not a fit. Owners of service firms often know this before anyone tells them.
  • Confidentiality and staff privacy. Rights to license mean owning the material, or being able to license it without breaching client confidentiality, employee privacy or third-party contracts.
  • Exclusivity. SourceX's licenses are exclusive for AI training for the selected data, though future data and unselected records are not included. Some owners will not give up that option for any price.
  • A file is not a dataset. Businesses that only want to sell a one-off file with no documentation or context are usually not a fit.
  • Distrust. The FTC staff post warns companies against widening their use of data quietly. An owner who has watched a vendor change its terms may distrust the whole subject, and that is a reasonable position.

An owner who stays on rung 1 has reached a complete outcome. For the ones who wonder whether licensing signals trouble, whether selling data means a company is failing and what changed since earlier data monetization attempts answer the two questions most likely to come next.

What trusted advisors and executives add

A headline cannot do two things an advisor can: describe the company's systems accurately and ask the person who decides. Accountants, consultants, managed service providers and network owners usually know which systems exist and who runs them, and they already have the owner's trust. Executives such as the COO, the CFO or the head of IT often own the records and can prepare the case; the CEO authorizes. The most useful thing an advisor does is help an owner move from rung 2 to rung 3 with accurate metadata, then step back. Collecting records, samples or logins is never part of it. The introduction economy guide maps which advisors see both the owner and the systems, and the AI data supply chain shows where the introduction sits among the other links.

The owner perspective interview series, designed as a panel

A single survey after a news cycle measures the news cycle. A panel interviews the same owners more than once, which is how a one-time spike separates from a lasting change.

Who. Owners or CEOs of companies with 50 or more full-time employees at peak, recruited with written consent from at least three partners' networks, about 30 per cohort, with a second cohort recruited six months later.

What is recorded at each wave. The highest rung reached; the trigger the owner names (news, a vendor's terms, an advisor, a peer, a buyer's outreach); the main hesitation in their own words, coded into the categories above; and whether they discussed it with anyone else at the company.

Who counts. Everyone invited is in the denominator, including owners who declined to be interviewed, and dropouts are reported wave by wave.

What is never collected. Records, samples or system access; quotes published without written approval; names in the report.

WaveTimingCore questionsWhat it separates
1Before any mention of the programHave you heard of companies licensing records for AI? From whom? What would you want to know first?Baseline awareness
2Within two weeks after a dated public signalDid you notice it? Did it change anything you did or asked?Response to a single event
390 days after wave 1What have you done since? Who at the company did you talk to?Lasting change

The outcomes and the decision rules are fixed before wave 1. SourceX has not supplied owner interviews, and the figures are not yet published.

Telling changed interest from a one-time campaign response

  • Changed interest. The share of panel owners at rung 3 or higher is greater at wave 3 than at wave 1, in both cohorts.
  • A campaign response. A rise at wave 2 that is gone by wave 3, or a rise only in the cohort recruited straight after a headline.
  • Not evidence either way. Clicks, webinar sign-ups and inbound questions to a partner, which measure attention rather than willingness.

A partner can keep a small version of the same log for their own conversations: the date of the first mention, the trigger named, the highest rung reached and where it stopped. It will not prove a trend, but it will show whether your own network is moving.

A practical next step for an owner who wants to explore

Start with the fit facts. The company's full-time headcount must have peaked at 50 or more, contractors aside, and it needs years of records created by its own operations, the right to license that history and a CEO who is keen to talk. Companies in the U.S., Canada and the EU qualify, as do companies in the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia when at least 70% of the data is in English. The company fit checker gives a preliminary, non-binding screen. Applying means answering a few questions with no files, and qualifying does not guarantee an offer. Some answers wait for SourceX's review however good the screen looks: whether the records match a current program, what the customer contracts permit and what any license would cover.

Limits of this reading

  • The signals come from the demand side. An AI developer, a regulator and a buyer are talking about business data. None of the three asked owners anything.
  • The OpenAI page is older. Its terms may have changed since it was published.
  • The FTC post is staff guidance. It is not a rule, and the excerpt we relied on is a paraphrase.
  • micro1's figure is a plan. It is a dated company announcement by a business that competes for the same records, not evidence of spending.
  • Demand moves. SourceX's criteria follow each program, so an owner whose company misses today's program could still match a later one.

Owners' beliefs before a first call are covered in what founders think data licensing means, and more guides sit in the guides hub.

Next step

Refer a company. No cap on rewards. Default reward on a company's first qualifying transaction: the greater of $50,000 plus 5% and 15% of Actual Supplier Payout, with no cap; further qualifying transactions first agreed within 12 months of the first Closing pay 15%. Eligibility and payment conditions apply, including the USD 100,000 Minimum Transaction, and a signed Schedule A or addendum can differ. A lead, meeting or signature alone does not trigger payment.

Once an owner in your network reaches rung 4 and the CEO asks to talk, your part is short: a tracked link via partner registration, a screen against the qualification rules and a careful read of the terms. An owner on rungs 1 to 3 needs an answer, not an introduction.

  1. Step 1Share your linkSend your personal link to a company you know.
  2. Step 2Company appliesThe company applies itself at /apply.
  3. Step 3Company signs and is paidThe company signs, delivers its data and is paid by SourceX under its agreement.
  4. Step 4You get your rewardthe greater of $50,000 plus 5% and 15% of the actual supplier payout, with no cap. A reward becomes payable after SourceX has paid the referred company under its signed agreement and the payout conditions in the program terms are met, including finance approval and your tax and payment details.

Common questions

Is it fair to say more owners want to license their data now?

Not yet. The verifiable signals show that AI developers, regulators and buyers are talking about business data; they say nothing about how many owners would license. Until a panel like the one on this page reports, describe what owners are hearing, not what they want.

An owner asked me about AI after a news story. Should I introduce them?

Not on the question alone. That is rung 1. Confirm the fit facts at metadata level, then ask whether the CEO wants a conversation with SourceX; only a yes from the CEO makes an introduction appropriate.

Does a reluctant owner hurt my standing as a partner?

No. A lead or meeting alone never earns a payable reward, so an owner who stays curious and goes no further costs you nothing in the program. Record the rung they reached and leave it there.

Sources

  1. openai.com/index/data-partnerships
  2. ftc.gov/policy/advocacy-research/tech-at-ftc/2024/02…
  3. micro1.ai/blog/micro1-commits-1b-to-enterprise-data-ac…
  4. refer.sourcex.si/who-qualifies
  5. sourcex.si/how-it-works
  6. refer.sourcex.si/terms

Free resources

Know a company with valuable proprietary data?

Become a referral partner, get your link and introduce the company’s CEO (in an owner-run company, usually the owner). Referral partners can join from most countries. We can't work with partners in countries or regions under U.S. sanctions or on U.S. restricted-party lists. The companies you introduce should be based in the U.S., Canada and the EU, or in the UK, Australia, New Zealand, India, Singapore, South Africa, Latin America, the UAE and Saudi Arabia with at least 70% of their data in English.

Refer a company →

I own a business

Explore licensing your company's data directly to SourceX for AI training and evaluation. Start a short assessment; no uploads needed.

Start an assessment