DataPacts

Education Companies Licensing Content to AI

Checked

At a glance

An education company can license some of its content to AI companies, but who owns the training right in a course or textbook, and whether student data sits inside the material, decide what can leave.

Independent reference. Every buyer is covered on the same terms, every figure carries its source and check date, and no buyer pays to be listed or placed higher.

Request an independent introduction

Tell us what data your company holds. We reply with the buyers that fit and what their terms leave out. Your details reach a buyer only after you agree to a named introduction.

IndependentYour details stay with us until you agreeNot selling is a valid outcome

Mercor, a buyer program, asks for content "the company holds the rights to license for AI training" and names textbook and courseware companies and e-learning platforms among its target industries (listing read 2026-11-17).

Education companies also speak to this in their own filings: Chegg states that it has licensed certain content to other entities, and Wiley reports Learning AI licensing revenue of USD 16.0 million for the year to 2026-04-30. No source checked gives a price per course or title. Education content is one sector in AI training data licensing.

DataPacts Editorial Team read the buyer pages, filings and rules cited here on 2026-11-17 or earlier as dated, following how we source, date and check every figure.

What Education Companies Hold That AI Buyers Ask For

Education companies hold four kinds of material that a buyer's page or an annual report touches: course and video content, textbooks and courseware, assessments and question banks, and learner and tutoring records.

The table compares the four company types, as of 2026-11-17; the buyer column is the buyer's own wording.

Company type What it holds Who usually decides licensing What a buyer page or filing says Source
Course provider or e-learning platform course videos, lecture recordings, learner submissions the instructor or employer for course content, the learner for own submissions, and the platform agreement between them Mercor industries include "e-learning platforms"; asks for "Course material" and "Video with transcripts" api.data.mercor.com listing publisher; coursera.org/about/terms
Publisher of learning material textbooks, courseware, question banks the publisher or author under the publishing contract Mercor industries include "textbook and courseware companies"; nice to have "STEM textbooks and lecture recordings" api.data.mercor.com listing publisher
Tutoring firm session recordings, chat transcripts, tutor-written explanations tutor contracts and the service terms, plus student data rules no program checked names tutoring transcripts n/a
Edtech software vendor usage logs, assessment results, learner profiles the school contract and education privacy law no program checked names these n/a

Mercor's publisher listing also asks for "Full texts", and its "not a fit" lines exclude "Content distributed under someone else's license" and content "already open access".

No program checked names assessments, question banks or tutoring transcripts, and Mercor's deal range for the listing is a whole-deal claim, shown in the filings table below. The data-type view sits under course and education content as AI training data.

Content Rights and Student Data Are Two Sets of Rules

A license to AI covers what the company made or owns, and the records of the students and children who used the product follow different rules that the company cannot sign away by owning the course.

Content the Company Made or Licensed In

Course videos, textbooks and question banks are the company's to license only to the extent it holds the training right, which a platform, publishing or distribution agreement can limit.

Mercor's listing names "Content distributed under someone else's license" as not a fit. Coursera's Terms of Use (effective 2026-01-01, read 2026-11-17) show a platform agreement deciding: content offerings are "governed by the relevant agreements in place between Coursera and Content Providers".

Third-party material inside slides and readings is not the company's to license. Open access and Creative Commons material carries its own terms and is not developed here.

Student and Child Data in Learning Products

Student work, tutoring chats, forum posts and usage logs identify learners, so the company licenses them only under the education privacy rules that attach to the person.

FERPA covers the education records of schools that receive funds under U.S. Department of Education programs (34 CFR 99.1), and an edtech vendor to which a school outsources a service may be treated as a school official under 34 CFR 99.31(a)(1)(i)(B), which ties it to 34 CFR 99.33(a): use only for the purposes of the disclosure, and no redisclosure without consent. All routes are set out under FERPA and licensing education records for AI.

Release without consent needs removal of all personal information and a reasonable determination that students are not identifiable "whether through single or multiple releases" (34 CFR 99.31(b)(1)).

Coursera's Terms of Use leave learners the intellectual property rights in their own submissions and give Coursera a "fully-transferable, royalty-free, perpetual, sublicensable, non-exclusive, worldwide license" in them. A license to the platform is a different thing from a license to a buyer, and this page does not read whether it covers AI.

The amended COPPA Rule (90 FR 16918, effective 2025-06-23, compliance date 2026-04-22) is explained at 90 FR 16950: disclosing a child's personal information to train AI is not integral to the service and needs separate verifiable parental consent under 16 CFR 312.5(a)(2). The consent rule is set out under COPPA and children's data in AI training.

Instructor, Tutor and Creator Contracts Decide Who Can Sign

The person who made a course or a lesson, not the company that sells it, holds the copyright unless the law or a signed writing says otherwise.

The table sets out who holds the training right in four situations, as of 2026-11-17.

Situation Who holds the training right Wording read Source
Employee instructor, course made within the job the employer, unless a signed writing says otherwise 17 U.S.C. 201(b); 101 "work prepared by an employee within the scope of employment" govinfo.gov
Freelance or guest instructor the instructor, unless a signed assignment or license says otherwise; commissioned works count as made for hire only in nine listed categories with a signed writing 17 U.S.C. 101 govinfo.gov
Tutor delivering sessions depends on the tutor's contract and the service terms no clause read n/a
Creator on a platform the agreement between the platform and the creator Coursera: offerings "governed by the relevant agreements in place between Coursera and Content Providers" coursera.org/about/terms

An instructor agreement is read by the company's own counsel for six points: the assignment or license of rights, the scope of the right granted, whether AI training or sublicensing is mentioned, moral or attribution terms, revenue share, and termination.

No education company's instructor agreement text was read for this page, so the list is a checklist, not a finding. A company runs a rights audit across every instructor and platform agreement before it describes a set to a buyer.

Recordings show instructors' faces and voices, which raises likeness questions that sit on the law pages below. Whether a tutor is an employee or a contractor is a classification question this page does not answer.

What Education Companies State in Their Own Filings About AI Licensing

Four education-side statements are on record as of 2026-11-17, from three annual reports and one buyer listing, and none of them gives a price per course or title.

The table lists the four statements in alphabetical order by company; each is the company's own, not a finding of DataPacts.

Company Document and date What it states Source
Chegg, Inc. Form 10-K for 2025, filed 2026-03-09 "we have licensed certain of our content to other entities who may use this content to train AI models that may compete with our products"; no counterparty, amount or terms sec.gov, accession 0001364954-26-000021
McGraw Hill, Inc. Form 10-K for the year to 2026-03-31, filed 2026-06-11 AI training on unlicensed content is "currently unsettled under the law"; a competitor or other data source "could make a deal to train generative AI models on their data"; the passage discloses no license of its own sec.gov, accession 0001951070-26-000022
Mercor (buyer program) Publisher or Education Company listing, read 2026-11-17 USD 35,000 to USD 370,000, "Depends on catalog size and rights"; 10 to 500 employees; the program's claim api.data.mercor.com listing publisher
Wiley Form 10-K for the year to 2026-04-30 Learning AI licensing revenue USD 16.0 million, against approximately USD 29 million a year earlier; Wiley's claim sec.gov, accession 0001628280-26-045111

A passage that discloses no license does not mean the company has none. Cengage Learning is a plaintiff in suits against Meta (N.D. Cal. 3:26-cv-11415-VC, filed 2026-05-05, pending as of 2026-11-09) and Google (S.D.N.Y. 1:26-cv-05870-LAP, filed 2026-07-10, pending as of 2026-11-17), and McGraw Hill LLC, named separately from the registrant above, is a plaintiff in the Meta suit.

Both concern training copies, not licenses. The sources read hold no AI licensing statement for Coursera, Duolingo, Khan Academy or Pearson, and not found does not mean no deal. Wiley's licensing record is set out under Wiley's AI licensing deals.

Benefits, Risks, Downsides and Alternatives for an Education Company

Licensing can earn income from content an education company has already produced and holds the rights to, but co-holders, platform agreements and student data limit what can go, and not licensing is a valid outcome.

Benefits

  • A named buyer asks for course material, textbooks and lecture video with transcripts (Mercor, read 2026-11-17).
  • Content already held in a CMS or LMS is the form that listing names.
  • Chegg's annual report shows that a content license exists on the record.

Risks

  • Breach of a platform or distribution agreement.
  • An instructor's or a third party's unlicensed material inside the set.
  • Student or child data inside the set, which FERPA and COPPA reach.
  • Competitors trained on the licensed content, a concern Chegg and McGraw Hill both state.

Downsides

  • No public price per course or title, and Mercor's range is a claim.
  • Rights clearing course by course.
  • A trained model keeps what it learned.

Alternatives

Where Education Licensing Meets Other Holders, the Law and the Seller Path

Education companies share course video with production studios, client work with agencies, and journals and monographs with academic publishers, and each page sets out its own limits. Education is one sector in data licensing by industry, next to video and photo production studios licensing footage and marketing, design and web agencies licensing data.

Journals and monographs sit with academic and scientific publishers licensing to AI, and institutions with universities and research groups licensing data. Textbooks follow the rules for books as AI training data, and lecture recordings those for selling video for AI training.

The full map is AI training data law by jurisdiction. Instructors on camera raise voice and likeness rights in AI training data, staff and tutor records fall under employee data and AI training, learner records under privacy law and AI training data, and the publisher suits against Meta and Google are tracked under AI copyright lawsuits over training data.

  1. Check data sale readiness.
  2. Review customer consent and confidentiality clauses.
  3. Remove learner data through de-identification before licensing data.
  4. Start choosing a buyer for your data.
  5. Negotiate the clauses of AI data licensing agreements.

This page is information, not legal advice.

Buyer and Law Pages Named on This Page

Each buyer program above has its own page with dates and sources.

Frequently Asked Questions

Can an edtech company license student data to AI companies?

Not by owning the product. FERPA and COPPA attach to the students and children, a vendor acting as a school official is limited to the purposes of the disclosure (34 CFR 99.33(a)), and disclosing a child's data to train AI needs separate parental consent (90 FR 16950).

Who can license a course an instructor made for an education company?

For an employee's course within the scope of the job, the employer owns the copyright (17 U.S.C. 201(b)). A freelance instructor's course needs a signed assignment or license, and a platform agreement can add limits.

How much do AI companies pay education companies for content?

No price per course or title is on record. Mercor states a whole-deal range of USD 35,000 to USD 370,000 for publishers and education companies, its own claim read 2026-11-17, and Wiley's revenue is a company total, not a price. See what company data is worth to AI companies.