European teachers are adopting AI rapidly, but want tools built for education

Sanoma Learning's survey of more than 20,000 teachers in 14 European countries found teacher AI use increased to 63% in 2026, while only 16% believe general-purpose AI improves learning outcomes. Between 75% and 93% of respondents want education-specific AI tools, and roughly two-thirds in major markets favor AI integrated into digital learning materials. The findings support demand for pedagogy-led educational AI and blended print-digital products, but highlight persistent concerns over general-purpose AI, workload and personalization constraints.

Analysis

The investable implication is not broad “AI-in-education” demand, but a shift in bargaining power toward incumbent curriculum owners with local content, school relationships and procurement approvals. Sanoma’s installed content base can turn AI features into retention and modest ARPU uplift rather than incur customer-acquisition costs; the near-term economic benefit is lower teacher-preparation friction and reduced churn, not a step-change in seat growth. This favors European curriculum incumbents such as Sanoma and Pearson (PSON.L) over horizontal AI vendors whose classroom monetization is constrained by safeguarding, data-residency and curriculum-validation requirements.

The print preference is a more important margin signal than an AI signal. A durable blended model protects high-margin proprietary content and limits full digital cannibalization, but also means AI investment will initially be incremental opex rather than a clean software-margin expansion. Sanoma’s survey is directionally useful but self-interested: it does not establish willingness to pay, school-budget availability, or procurement conversion. Without evidence of price realization or renewal improvement, the release alone is insufficient to underwrite earnings revisions.

Over the next 1-3 months, watch whether Sanoma quantifies AI attachment rates, paid pilots, renewal rates, or content-development productivity at its next results update; those are the catalysts that can move consensus EBITDA expectations. Over 6-18 months, a credible education-specific AI layer could widen the moat against generic models, while pressuring smaller publishers that lack the capital and localized datasets to build compliant tools. The thesis is falsified if AI features are bundled free to defend renewals, producing rising technology spend without measurable net revenue retention or gross-margin expansion.

Key Decisions for Investors

• No immediate directional trade: impact is low and the underlying evidence is a company-sponsored survey rather than monetization data. Place SANOMA/SAA1V on watch for the next earnings release; upgrade only if management discloses paid AI penetration or raises Learning organic-growth/EBITA guidance.
• Construct a 6-12 month relative-value watchlist: long SANOMA (or SAA1V) versus short a smaller education-content peer with weaker local-curriculum exposure, only after validating valuation, liquidity and AI investment disclosures. The intended payoff is multiple divergence from recurring-content moat expansion; exit if Sanoma’s Learning margin falls despite stable revenue.
• For liquid exposure, monitor PSON.L as the cleaner listed education-AI proxy. A long entry is justified only on evidence that AI is driving higher digital-courseware renewal or pricing rather than promotional bundling; use a 10-15% downside stop or a guidance-cut trigger, since public-sector education budgets can delay conversion by 1-2 procurement cycles.
• Avoid NIQ as an AI-education read-through: its role is data collection, not a direct beneficiary of educational AI adoption. Treat any price response as unrelated unless a separate commercial data-product catalyst emerges.