DataPacts

Cohere: Data Buyer Profile, Program Terms, Payouts and Reputation

Checked

At a glance

Cohere, the Canadian enterprise model developer named in court as Cohere Inc., has no public data licensing deal and no program for companies that want to license data to it in the sources checked as of 2026-11-09. Its record as a data acquirer is a lawsuit by fourteen news publishers over training and retrieval, which was in discovery as of the latest docket entry on record, 2026-11-02.

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

Every figure on this page is a dated statement by a court filing, Cohere or a regulator page. The profile follows the same method as every buyer profile in this independent reference on AI training data licensing.

Each number carries its source and check date, and how DataPacts sources and dates every figure is set out in the methodology.

What Cohere Acquires on Record: No Deals and No Seller Program

No acquisition route for training data is documented for Cohere in the sources checked as of 2026-11-09: no content license, no dataset purchase and no seller program.

  • Content licenses: none on record.
  • Company data program: none on record.
  • Web crawling: no Cohere crawler in the record.
  • Training sources as alleged: the publishers' complaint, an allegation and not a finding.

What is absent matters as much as what is present. Cohere has no statement in the record about what its models train on, no dataset purchase is on record, and no terms for contributors or sellers appear in the sources checked.

For a seller the gap has a practical effect. When a lab discloses neither a license nor a source, the difference between licensed versus scraped training data cannot be read from the lab's own pages, and the holder has only the court record and regulator pages to go on.

Cohere Company Identity and Contracting Entity

The legal entity on record for Cohere is Cohere Inc., the name under which it is sued in the Southern District of New York, with Canada as its country.

The table lists the identity attributes on record, as of 2026-11-09.

Attribute Value Source
Legal name Cohere Inc., as named in the S.D.N.Y. caption Complaint, 1:25-cv-01305
Corporate registry number not checked n/a
Country Canada cohere.com
Website cohere.com cohere.com
Type frontier lab, enterprise LLM developer (Command) cohere.com
Founding date not recorded n/a
Founders, chief executive, headcount, address not recorded in the sources checked n/a

The legal name comes from the court caption and the country from the record, while the Canadian corporate registry number was not checked and none is given here.

The sources checked do not state a founding date, founders, a chief executive, headcount or an address. The page on rules for licensing data for AI training in Canada covers the Canadian frame that applies to a Canadian licensee.

Cohere Program Terms and Payouts, Dated

Cohere publishes no program, application page or terms for licensing company data in the sources checked; the table lists each term a data holder would check and what the record holds, as of 2026-11-09.

Term What the record shows Checked Source
Company data program or application page none on record 2026-11-09 cohere.com
License scope, term, exclusivity and deletion for company data none published 2026-11-09 cohere.com
Payout to data holders none stated 2026-11-09 cohere.com
Referral fee none recorded 2026-11-09 cohere.com
Content or data licenses with any counterparty none on record 2026-11-09 cohere.com
Crawler and robots.txt opt-out no Cohere crawler recorded in the sources checked 2026-11-09 n/a
EU training content summary none recorded in the sources checked 2026-11-09 n/a
GPAI Code of Practice listed signatory (Commission page updated 2026-10-07) 2026-11-09 European Commission

Every company-data row reads "none on record". That leaves no scope, term, exclusivity, deletion or payment trigger to compare against the program terms compared across buyers that do publish them, and the wording does not say that Cohere never acquires data.

The only binding party a licensor could identify is the entity named above, so any terms would be negotiated from zero.

Cohere Reputation: The Court Record as of 2026-11-02

Fourteen news publishers led by Advance Local Media sued Cohere Inc. in the Southern District of New York on 2025-02-13, and the case was in discovery as of the latest docket entry on record, 2026-11-02.

The table gives the court record as of 2026-11-02, with values taken from court filings.

Field Value Source
Case Advance Local Media LLC et al. v. Cohere Inc. Complaint
Court and docket U.S. District Court, S.D.N.Y., 1:25-cv-01305-CM-OTW Complaint
Filed 2025-02-13 Complaint, Doc. 1
Plaintiffs Advance Local Media, Conde Nast, The Atlantic, Forbes, Guardian News & Media, Insider, Los Angeles Times, McClatchy, Newsday, Plain Dealer, Politico, The Republican, Toronto Star, Vox Media Complaint
Claims (allegations) copyright infringement in training and retrieval-augmented outputs; trademark claims Complaint
Ruling partial motion to dismiss denied 2025-11-13 (Doc. 59) Doc. 59
Discovery plaintiffs' production of approximately 43,000 documents (parties' joint letter, 2026-09-25, Doc. 149) Doc. 149
Latest orders 2026-11-02: Doc. 168 denies Cohere's letter motion to compel (ECF 161, 162); Doc. 167 concerns a clawback Doc. 168
Status discovery; no settlement on record Doc. 168

The plaintiffs allege that Cohere copied their articles to train its models and that its retrieval-augmented outputs reproduce them, and they add trademark claims. These are allegations in a complaint and not findings.

On the docket, the court denied Cohere's partial motion to dismiss on 2025-11-13 (Doc. 59). The parties' joint letter of 2026-09-25 states that the plaintiffs had produced approximately 43,000 documents, which is the parties' own figure.

On 2026-11-02 the court denied Cohere's letter motion to compel (Doc. 168) because Cohere had agreed not to seek that discovery, and issued a separate clawback order (Doc. 167). The full Advance Local Media v. Cohere case record has its own page. The docket was not re-checked after 2026-11-02.

For a licensor, no settlement or license with the plaintiffs is on record. No other court case, breach notice or regulator action involving Cohere was found in the sources checked as of 2026-11-09.

Cohere Training Data Statements, Crawlers and EU Obligations

Cohere's own account of its training data is not in the record checked here, and neither is an EU training content summary, but the European Commission lists Cohere as a signatory of the GPAI Code of Practice.

The complaint's allegations are not a Cohere statement, so none is summarized as one.

No Cohere crawler name, user agent or robots.txt opt-out is recorded in the sources checked as of 2026-11-09, so this page names none. Site owners manage crawlers in general through robots.txt, and the page on AI crawlers explains how to allow, block or charge them.

Cohere is listed among the 22 signatories of the voluntary GPAI Code of Practice (Commission page last updated 2026-10-07, checked 2026-11-09). The Code was published on 2025-07-10 and is not binding law. Its Copyright chapter has signatories' crawlers follow robots.txt under IETF RFC 9309 (Measure 1.3).

No training content summary by Cohere under Article 53(1)(d) of Regulation (EU) 2024/1689 is recorded in the sources checked.

The GPAI obligations apply from 2025-08-02, Commission enforcement from 2026-08-02, and providers of models placed on the market before 2025-08-02 have until 2027-08-02. In the Commission's template, licensed datasets appear only as Yes or No plus modality.

The Article 53 training data summary and copyright policy page holds the template and the full rules.

Benefits, Risks, Downsides and Alternatives for a Data Holder Considering Cohere

For a data holder, Cohere offers an identifiable Canadian entity but no published route, terms or precedent license, so the practical alternatives are buyers that publish programs, and the option of not licensing at all.

Benefits (as the record supports them):

  • An identifiable counterparty named in a federal court caption, Cohere Inc.
  • A listed GPAI Code signatory, with a robots.txt commitment in the Copyright chapter.

Risks:

  • Pending litigation over training and retrieval on news content, so a licensor's data would enter a contested setting.
  • No published license terms, so scope, exclusivity, deletion and payment start from zero.
  • No training content summary recorded, so the reporting of licensed data cannot be checked.

Downsides:

  • No application route, payout or referral fee for company data is on record.
  • No precedent license with any data holder exists to compare against.

Alternatives (alphabetical; each one attribute as the party's claim, checked 2026-11-09):

  • Handshake AI Data Partnerships: Handshake states "Earn $100K to $4M+", while the estimator on the same page reads "Your data could be worth up to $100,000".
  • Labelbox Prism: datasets "typically start around $50K", and enterprise partnerships "can reach $1M+" (2026-11-09).
  • micro1 Company Data Partnerships: USD 100K to 500K for qualified enterprise data partnerships (tier 1, 2026-11-09).
  • Microsoft Publisher Content Marketplace, for publishers: publishers "will be paid on delivered value", with no rates published.
  • Scale Data Partnerships: "$10K-$1M+ illustrative value per data partnership" (2026-11-09).
  • Not licensing at all, which keeps the data and every later option open.

These figures are each party's claims, not payments made. The wider set of alternatives to selling data includes keeping the data.

Cohere Among Frontier Labs That Acquire Data

Cohere is one of two frontier labs on the labs overview with no deal or seller program on record, and one of several labs facing news publishers in court.

It sits in the comparison of frontier AI labs as data buyers, and the hub on who buys AI training data places it beside every other buyer. When a lab has no company program, choosing a buyer for your data covers the next step.

More About Cohere as a Data Buyer

Where Cohere's Court and Licensing Records Are Tracked

Each record above has its own tracker page with full filings, sources and status.

Frequently Asked Questions About Cohere and Training Data

Does Cohere pay for training data?

No payment for training data by Cohere is on record: no licensing deal, dataset purchase or seller program appears in the sources checked as of 2026-11-09.

Can a company license its data to Cohere?

No Cohere program, application page or terms for company records appear in the sources checked as of 2026-11-09.

Who is suing Cohere over training data?

Fourteen news publishers, led by Advance Local Media, in S.D.N.Y. case 1:25-cv-01305-CM-OTW filed 2025-02-13; the case was in discovery as of 2026-11-02.

Cohere Inc., the defendant named in the federal court caption, with Canada as its country; its Canadian registry number is not recorded here.

What does Cohere do?

Cohere develops large language models for enterprises, including its Command models.