Arabic & GCC data · now in early access
License your data to AI.
Get cited by it.
We help organizations in the Gulf turn proprietary Arabic data into new revenue while supplying the AI models and agents that need it.
- 312
- licensing deals tracked, Jan 2023 – Sep 2026
- 3 → 39
- deals per quarter over the period
- $9.4M → $3.6M
- median annual deal value
- 170 : 441
- Gulf vs Modern Standard Arabic entries in the Arabic AI Atlas
Sources: AI Industry Reviews (Oct 2026; one secondary tracker) and Middle East AI News. Directional figures.
Your organization already holds data AI developers can't find
Archives, recordings and knowledge your teams produced over years. Buyers are signing many smaller deals with regional licensors, and paying more by usage. A small holder can't run each negotiation alone.
Datasets AI developers are looking for
Examples of the Arabic and GCC data we package and license.
Arabic media & publishing
Legal & regulatory
Real estate & tourism
Product & technical
Conversational Arabic
Built with privacy and control first
Every dataset ships with a datasheet and provenance record. Before any data moves, we agree the boundaries with you.
Datasheet · illustrative
- Content
- Transcribed speech, Gulf Arabic
- Language / dialect
- Arabic (Gulf)
- Collection method
- Documented, with speaker consent
- Rights
- Seller-warranted, indemnified
- Personal data
- De-identified; sampling audit passed
- Permitted use
- Train · Ground · Evaluate (per license)
- Known gaps
- Stated in the datasheet
How it works
A simple, privacy-first process designed for long-term partnership.
- 1
Apply
Tell us what you hold. A short form is enough.
- 2
Evaluate
We review a sample and decide with you whether it is licensable.
- 3
Agree
We set scope, rights, price basis and permitted uses in a license.
- 4
Deliver
Buyers get an expiring download link or query access through the API. Sale proceeds are held for 14 days before a seller can withdraw them.
One dataset, three licenses
Price follows use, not file size.
Train
Versioned snapshot
Ground
Agent endpoint
Evaluate
Curated test set
A marketplace, not a single-buyer program
Many buyers, one agreement
Arabic and Gulf first
Citations, not just a payout
Frequently asked questions
What kinds of data can I list?
Arabic text and transcribed speech to start: news and publishing archives, legal and regulatory texts, catalogs, manuals and similar non-personal knowledge. Conversational data comes once consent and de-identification are proven.
Do you accept health or children's data?
No. Special-category personal data is out of scope in the first release.
Who owns the data after licensing?
You do. Buyers receive a license for a stated use, term and geography; resale and re-identification are prohibited.
How is pricing set?
By use (train, ground, evaluate), exclusivity, metadata richness and whether the content is already public. We agree the basis with you before any listing.
Is this live?
We are in early access, running hands-on pilot licenses. The self-serve platform follows.