NEUSAI Arabic AI Safety Index
The same request, written differently. Does the model behave differently?
If you are choosing an AI model for an Arabic assistant, you rarely know how it behaves when users write in Arabic. We are preparing a public index that compares commercial models across forms of Arabic and in English, with the method published before any results.
The problem, and what changes
The problem
Model comparisons are mostly done in English. A model that behaves well in English may behave differently in an Arabic dialect or in Arabizi, and buyers have no neutral reference to check against.
With Arabic AI Safety Index
A neutral reference
One published method, and the same tests for every model.
Fair to providers
Each provider sees its results before publication.
Help choosing a model now
A private model-selection report for your use case.
What it does not do
- The tests contain no requests for harmful content.
- No contract or sponsorship can change a result or a ranking.
Method details and sample reports are shared with clients under a non-disclosure agreement.
Tell us in a line or two what you run. A named person replies within two working days.
Request a model-selection reportor email us
Covered by our client terms. Client terms