Linear's terms, effective June 9, 2026, state that the customer owns its User Submissions, and its AI Addendum, effective November 13, 2026, states that Linear will not use Customer Data to train, fine-tune or otherwise improve its own AI models.
Linear is one source system in AI training data licensing, and six programs name it among the records they want (checked between November 9 and November 17, 2026). The DataPacts Editorial Team dates each term and limit below; the method is explained in how we source, date and check every figure.
What a Linear Workspace Holds and Why Buyers Name It
A Linear workspace holds issues with their fields, projects, cycles, initiatives and customer requests, and six programs name Linear among the systems whose records they want.
The workspace CSV carries fields such as ID, team, title, description, status, estimate, priority, project, creator, assignee, labels, cycle details, dates from created to archived, due date, parent issue, initiatives, project milestone and SLA status (read November 17, 2026). Customer requests and projects sit in separate exports.
The value of ticket histories as a data family, and the Jira counterpart, are covered elsewhere: see selling issue trackers and project histories for AI training.
Who Controls Linear Data in a License
Four layers decide what Linear data a company can license: its own terms with Linear, Linear's AI commitments, the data protection terms for personal data, and the people named in the issues.
Table 1 compares the four layers (as of 2026-11-17). The service is provided by Linear Orbit, Inc., and no registry number is published.
| Layer | Who | What the record shows | Effect on a license | Source |
|---|---|---|---|---|
| Customer terms | the licensing company and Linear Orbit, Inc. | terms effective 2026-06-09: customer owns User Submissions; Linear license limited to providing the service; Linear owns aggregated, anonymized Service Data | the company controls its issues; exclusivity wording is a question for counsel | linear.app/terms |
| AI commitments | Linear | AI Addendum effective 2026-11-13: no training or improvement on Customer Data; providers barred from training | covers Linear's own use, not a buyer's | linear.app/legal/ai-addendum |
| Personal data terms | Linear as processor | DPA s. 2.1 processor role; s. 2.4 return or delete; s. 3 sub-processors with notice; Exhibit B lists AI sub-processors | personal data in issues stays subject to data protection law | linear.app/dpa |
| People in the issues | staff, customers, contractors | names, requests and free text | notice, consent or removal before delivery | see employee consent below |
Customer Terms: Ownership and Linear's Limited License
Linear's terms, effective June 9, 2026, state that the customer owns all right, title and interest in its User Submissions, and give Linear a non-exclusive license to use them to provide the service.
Section 10.2 describes that license as "non-exclusive, worldwide, royalty-free" and limited to providing the service. Section 10.3 says Linear owns Service Data, meaning performance and operation measures, when aggregated and anonymized. Section 10.5 sends AI use to the AI Addendum, which controls in a conflict.
Whether Linear's license or its Service Data clause meets a buyer's no-prior-grant or exclusivity term is a question for counsel and the buyer, and exclusive vs non-exclusive data licenses explains how such terms work. The terms text read does not describe a customer export right; the export routes sit in Linear's documentation.
The AI Addendum and the DPA
Linear's AI Services Addendum, effective November 13, 2026, states that Linear will not use Customer Data to train, fine-tune or otherwise improve its own AI models, and that the customer keeps ownership of its Customer Data.
Customer Data includes prompts and outputs (s. 13.1). Linear also requires its AI providers not to use Customer Data to train their own models and to process it with zero data retention where commercially available (s. 13.3), and it does not re-enable an AI feature a customer's administrator has switched off (s. 13.4).
In the Data Processing Addendum, Linear is the processor (s. 2.1), returns or deletes personal data at the customer's choice (s. 2.4), and lists AI sub-processors in Exhibit B, for example Anthropic, PBC (read November 17, 2026). These terms cover what Linear does, not what a buyer may do with an export.
How Linear Data Leaves the Workspace
Linear data reaches a buyer by one of four routes: a CSV export the company runs, a Markdown copy, the GraphQL API with webhooks, or a read-only connection or snapshot a program runs.
Table 2 compares the routes (as of 2026-11-17).
| Route | Who may run it | What it carries | Limits | Source |
|---|---|---|---|---|
| Workspace issue CSV | admins; owners only on Enterprise | issue fields; private teams by toggle | emailed link, expires after 12 hours; recorded in audit log | linear.app/docs/exporting-data |
| View issue CSV | members up to 250 issues; admins (owners on Enterprise) up to 2,000; guests cannot | same fields as the workspace export | per view | linear.app/docs/exporting-data |
| Project and initiative CSV | members and admins | project and initiative fields | up to 200 at a time | linear.app/docs/exporting-data |
| Customer request CSV | command menu, scoped to an issue, project or customer | customer requests | none stated | linear.app/docs/exporting-data |
| Markdown copy | a user copying an issue | title, description, comments and customer requests | per selection | linear.app/docs/exporting-data |
| GraphQL API and webhooks | API key or OAuth app | not stated | API key 2,500 requests and 3,000,000 points per hour; OAuth app 5,000 requests and 2,000,000 points per hour; default page 50 records | linear.app/developers/rate-limiting |
| Sync integrations | admins setting up Airbyte or Google Sheets | reporting copies; Sheets covers public teams only | not stated | linear.app/docs/exporting-data |
| Backup | none described on the export page | n/a | n/a | linear.app/docs/exporting-data |
| Read-only connection or snapshot | program (Scale AI, Troveo, Avelence) | anonymized inside the company's environment (Scale AI); no OAuth access or install (Troveo); separate copy (Avelence) | each program's own process | scale.com, troveo.ai, avelence.com |
Workspace CSV, View CSV and Who May Run Them
Admins can export every issue as a workspace CSV from Settings, Administration, Import Export; on Enterprise plans only owners can. The export is recorded in the audit log, and the emailed link expires after 12 hours.
- A view CSV covers up to 250 issues for members and up to 2,000 for admins, or owners on Enterprise.
- Guests cannot export issues, and project or initiative CSVs cover up to 200 at a time.
- The member list is a separate CSV of staff data, run by admins or Enterprise owners.
A company with more than 2,000 issues in a view uses the workspace export or the API.
What the CSV Leaves Out: Comments, Attachments, History
The CSV exports omit attachment files, their field lists do not include comments, and the export page lists no history or activity field (read November 17, 2026).
- Links to attachments may appear in issue descriptions, but the files are not exported.
- A Markdown copy of an issue captures title, description, comments and customer requests, one selection at a time.
- The export page describes no backup option.
A buyer asking for "everything in Linear" receives what the chosen route holds, so the manifest lists the gaps.
GraphQL API, Webhooks and Sync Integrations
Linear exposes data through a GraphQL API and webhooks, with request and complexity limits that cap how fast a bulk extraction can run.
An API key allows up to 2,500 requests and 3,000,000 complexity points per hour, an OAuth app up to 5,000 requests and 2,000,000 points, and the default page returns up to 50 records (read November 17, 2026).
Airbyte and Google Sheets sync data for reporting, and Sheets covers public teams only. Whether the API returns complete comment and history data is not recorded here, and third-party exporters and migration tools are not reviewed.
Read-Only Connections and Snapshots Offered by Programs
Some programs read or prepare the records themselves instead of receiving a raw CSV, and their published process decides who sees identifiable data first. For the wider comparison, see buyer connectors vs native exports.
Scale AI states that it anonymizes inside the company's environment before extraction (checked November 17, 2026).
Troveo, a marketplace, states that no OAuth access or install is required (checked November 17, 2026), and Avelence, a broker, states that it has no default system access and works on a separate copy (checked November 9, 2026). Troveo and Avelence handle data on the seller's behalf and do not buy it.
What to Remove From Linear Issues Before Delivery
Linear issues carry personal data, customer names and secrets in free text, so removal starts with identity fields and ends with descriptions, customer requests and attachment links. Methods for PII removal and de-identification apply to issue text as to any other text.
The removal list for a Linear export covers:
- creator and assignee names and the member list
- customer requests and the customers named in them
- credentials, tokens and URLs pasted into descriptions
- links to attachments and to other systems
- free-text names of colleagues, clients and contractors
Scale AI anonymization inside the company's environment is one published method, and Turing states that "De-identification and preparation are on us" (checked November 17, 2026). Customer-side confidentiality terms for customer requests sit in the company's own contracts. Staff named in issues are covered by employee notice and consent.
What Buyers and Intermediaries Ask For in Linear Records
Six programs name Linear among the records they want, and each publishes different terms, checked between November 9 and November 17, 2026.
Table 3 compares them alphabetically (as of 2026-11-17); every value is the program's own statement.
| Program | Role | Linear-related records named | Terms that matter for Linear data | Source |
|---|---|---|---|---|
| Avelence seller advisory | broker (intermediary) | "Jira/Linear/Asana/Notion" | no default system access; extraction and de-identification on a separate copy (checked 2026-11-09) | avelence.com |
| DataVendor (HUD) | marketplace (intermediary) | issue or ticket history (Linear, Jira) listed next to codebases | vendors set the price; platform fee 20 % of earnings at withdrawal (checked 2026-11-09) | docs.datavendor.ai |
| Mercor | data buyer | Linear among tools in the software-company listing (GitHub, GitLab, Bitbucket, Jira, Linear, Confluence) | listing range USD 50,000 to 500,000, more for large codebases or exclusive terms; published 2026-10-17 (checked 2026-11-09) | api.data.mercor.com |
| Scale Data Partnerships | human-data platform | "Jira, Confluence, Asana, Notion, Linear" | anonymizes inside the company's environment; per-package approval gates; exclusivity, term, deletion and retraining not published (checked 2026-11-17) | scale.com |
| Troveo business data | marketplace (intermediary; licenses on the seller's behalf) | "Jira, Linear, Asana, Notion, Figma" | seller retains ownership; no OAuth access or install required; exclusivity not stated (checked 2026-11-17) | troveo.ai |
| Turing Project Lazarus | human-data platform | work trackers including Jira and Linear (list not quoted) | exclusive license to Turing for the first 6 months, during which Turing may license the same record to more than one lab; Turing covers de-identification and preparation (checked 2026-11-17) | lazarus.turing.com |
No program publishes a price for Linear records. Mercor's range of USD 50,000 to 500,000, published October 17, 2026, covers its software-company listing with Linear among the named tools and is Mercor's claim; no seller-confirmed amount for Linear records is on the public record.
The full set of AI data partnership payouts and program terms compared sets these claims beside other programs.
Avelence, DataVendor and Troveo handle, list or place data and are not the licensee of last resort. The table lists programs by name and implies no ranking or endorsement.
Licensing Linear History: Benefits, Risks, Downsides and Alternatives
Licensing Linear history can earn income from records a company already keeps, and it can also expose staff, customers and secrets, so not licensing remains a valid result.
Benefits
- Income from records the company already keeps, as Mercor's listing range claims.
- Some programs carry preparation costs, as Turing states for de-identification.
- A read-only or in-environment route keeps raw data in place, as Scale AI and Troveo describe.
Risks
- Personal data and employee rules for staff named in issues.
- Credentials and secrets pasted into descriptions.
- An exclusive window blocks other buyers, as Turing's first 6 months do.
Downsides
- The CSV omits comments and attachment files.
- View limits and the Enterprise owner-only rule slow a large export.
- No published price exists for Linear records alone.
Alternatives
- License a subset: one team, closed issues or older years.
- License issue history together with the codebase, see selling or licensing source code and codebases.
- Keep the data and use Linear's administrator controls for its AI features.
- Do not license at all.
Linear Among the Work Trackers and the Seller Steps
Linear is one of the work trackers whose terms and export routes decide what a company can license, next to Jira, Asana and Notion. The parent page gives export rights and mechanics by source system.
Sibling pages cover Jira data export rights, Asana project export, Notion workspace export and Confluence data export, with Monday.com board export and Bitbucket export to follow. The seller steps that touch Linear run in this order:
- Run a rights audit of the workspaces.
- Build the data inventory of teams, years and size.
- Check customer consent and confidentiality clauses.
- Apply employee data and AI training rules by jurisdiction.
- Write the data manifest, including the CSV gaps.
- Plan delivering licensed data.
- Decide should you sell or license your data at all.
Linear Data After a Subscription Ends or a Closure
A company that is closing should export its Linear data while the workspace is active, because Linear's terms set deletion within 30 days after a termination for cause, on request or after workspace deletion, and keep Free Version data (s. 5.3, effective June 9, 2026). The steps are in selling data from a defunct company.
Deadlines at vendors are covered in exporting company data before shutting down, and the DPA lets the customer choose return or deletion.
Buyer Programs Cited on This Page
The program terms above come from each program's own pages, checked between November 9 and November 17, 2026. Programs cited, alphabetically:
- Avelence data buyer profile
- DataVendor (HUD) data buyer profile
- Mercor data buyer profile
- Scale AI data buyer profile
- Troveo data buyer profile
- Turing Project Lazarus data buyer profile
See also company data partnership programs and who buys AI training data.
Frequently Asked Questions
How do I export all issues from Linear?
An admin can export every issue as a workspace CSV, and on Enterprise plans only an owner can. The CSV leaves out attachment files and comments (read November 17, 2026).
Does Linear use customer data to train AI?
Linear's AI Addendum, effective November 13, 2026, states that Linear will not use Customer Data to train, fine-tune or otherwise improve its own AI models. Customers keep ownership of that data. The platform-wide how-to is in opting out of AI training.
Can a company license its Linear issues to AI companies?
A company can license Linear issues it controls once personal data, customer terms and secrets are dealt with, and several programs ask for them. Six programs name Linear in Table 3. The limits are in what company data cannot be licensed.
Does the Linear CSV export include comments?
No; the CSV field lists do not include comments, though a Markdown copy of an issue does. Attachment files are also excluded.