genAssess - AI Literacy Assessment (Core-6)

What it is

genAssess is a psychometric AI-literacy test for workforces, operated by Eunomia Ltd (company no. 16995212 per the vendor site, re-read 2026-10-05). The current product is Core-6: 48 questions across six competencies (AI Foundations, Application, Risk, Data, Evaluation, Oversight), 15-20 minutes, mobile-friendly, delivered through the Sova Assessment platform (all per the vendor site). The vendor states it was screened for differential item functioning across sex, age and ethnicity, and positions it as supporting the AI-literacy duty in EU AI Act Article 4.

Validation claim (vendor-reported)

The vendor reports a convergent-validity correlation of r = 0.83 (p < 0.001; sample n = 289; both figures stated on the vendor site, re-read 2026-10-05) against GLAT, the Generative AI Literacy Assessment Test (a published 20-item instrument; arXiv 2411.00283, validated on 355 higher-education students). The GLAT paper is real; the genAssess-vs-GLAT correlation is the vendor’s own claim and has no independent publication that we could find.

Access

No public pricing; enterprise pilots and licensing by request.

Context and comparison

• Earlier versions of this note described genAssess as a task-mapping “global benchmark for AI readiness” with peer benchmarking; the current site describes a literacy test. Task mapping and global benchmarking features are not confirmed on the current site and were removed.
• EU AI Act Art. 4 (AI literacy) has applied since 2 Feb 2025; the 2026 Digital Omnibus (Regulation (EU) 2026/1744, in force 2026-07-27) softened it to an obligation to take measures to support AI literacy (see the verified EU AI Act note).
• Versus paice-work: genAssess tests knowledge (what people know about AI); PAICE observes behaviour in an AI interaction (0-1000 score). They are complementary (opinion). workera-ai is a broader skills-intelligence platform.
• Other AI-literacy instruments mentioned in the earlier note (OECD AI Literacy Framework, SAIL4ALL, an attributed Mollick “~10 hours” rule of thumb) were not re-verified here and are omitted.

Open items

• Independent validation of the r = 0.83 GLAT claim: not found, vendor claim only (the figures themselves are confirmed on the vendor site).
• Pricing and customers: not public.

Related

paice-work, workera-ai, ai-learning

Sources

• https://genassess.com (accessed 2026-10-02 and 2026-10-05)
• https://arxiv.org/abs/2411.00283 (GLAT, via search results, accessed 2026-10-02)
• https://eu-nomia.com/ (search result, accessed 2026-10-02)