Our editorial method
Updated 15 October 2026
What we publish
LearnIA offers methods, examples and activities for using AI and checking its results. Workshops state what is simulated and what is actually executed. Resources are available without an account.
Sources and dates
Product information is checked against publisher documentation and announcements. Articles identify sources and relevant dates. An update date should reflect an actual content revision.
Three kinds of evidence
We distinguish publisher claims, third-party evaluations and tests actually performed by LearnIA. External scores retain their dates, metrics and known conditions. They are not LearnIA test results.
Reproducible comparisons
A LearnIA test should document tasks, data, versions, settings, tools, criteria and limitations, including outputs, errors and rework. A small sample does not establish general model superiority.
AI assistance and editorial checks
AI tools may help draft, translate or structure content. References, calculations and claims must be checked before publication. Invented examples are labelled fictional and are not customer testimonials.
Activities and progress
Activities include criteria or explained answers. Progress and answers may be stored in your browser. A completion record is self-reported, not a qualification or independent assessment.
Commercial links
Affiliate links are disclosed. A commission does not replace a comparison method. Personalised services are presented separately from free articles and resources.
Catalogue limitations
Not every historical article was checked line by line in the October 2026 release. Check current sources before buying a product. Four long courses currently have French-only lesson bodies, disclosed on their English views.
Request a correction
Send the URL, the relevant passage and a source supporting the correction.
Contact Dorian