The Public Language Index was built to answer a question that nobody was measuring: how is AI changing the way European institutions write?
Institutional communication carries cultural weight. The way a university writes reflects its intellectual tradition. The way a government ministry writes reflects its relationship with citizens. The way a bank writes reflects its understanding of trust. These are not aesthetic preferences. They are the accumulated expression of institutional identity, shaped by history, sector norms, regulatory context, and the linguistic traditions of the countries and languages in which they operate.
AI writing tools trained predominantly on English-language data are producing institutional content across Europe. The result is measurable: cultural markers are declining, idiomatic expression is thinning, vocabulary is narrowing, and organisations with radically different histories are beginning to produce linguistically identical communications. The Observatory has documented cases where pairwise similarity between unrelated organisations rose from 18% to 100% in under five years.
This is not an argument against AI. It is an argument for measurement. Organisations that understand how their language is changing can make informed decisions about where AI supports their voice and where it overwrites it. The PLI provides the evidence base for those decisions.
CopyMama S.R.L. is a linguistic intelligence consultancy based in Milan. Our mission is to protect, govern and teach what it means to write like humans. The company operates through three integrated pillars.
The research engine. Monitors how communication evolves in AI-influenced environments and publishes the Public Language Index. Houses PRISM™, the diagnostics engine behind every measurement.
The professional services unit. Builds verbal identity systems for institutions using VOICE™, a governance framework that translates Observatory diagnostics into editorial standards, tone guidelines and communication architecture.
The training pillar. Certifies professionals in expressive governance: the skills required to manage clarity, tone and identity across communication systems shaped by automation.
The Observatory's research feeds into CopyMama's other two pillars. PRISM™ diagnostics inform VOICE™ governance engagements in the Studio. Observatory benchmarks shape the training curriculum delivered through the Academy. Findings from professional practice flow back into the research programme. The three units form a cycle: what the Observatory discovers, the Studio applies, and the Academy teaches.
Mission: To protect, govern and teach what it means to write like humans.
The Observatory is designed around European research, regulatory and cultural priorities. All infrastructure is hosted within the European Union, and the platform's mission aligns with specific EU programmes and policy objectives.
The PLI archives pre-AI communication patterns across European languages and tracks where AI tools trained on predominantly English-language data are reshaping institutional expression in other linguistic traditions. Coverage currently spans ten languages, with expansion toward all 24 EU official languages underway through academic partnerships. This work aligns directly with Horizon Europe's cultural heritage and linguistic diversity objectives.
PRISM™ is classified under the EU AI Act's transparency risk tier, the only CopyMama system above minimal risk, because it evaluates content shaped by automated systems. Every measurement uses peer-reviewed statistical methods, produces confidence intervals, and is reproducible from source data. Every datapoint and analysis result carries a cryptographic blockchain timestamp. Full application of the EU AI Act begins August 2026.
The 2015–2019 baseline corpus preserves how European institutions actually wrote before generative AI reshaped public language. These records become more valuable over time: timestamped proof of authentic human communication patterns recorded before widespread AI adoption. The Observatory treats this archive as cultural infrastructure.
The Observatory's corpus, analysis pipeline and research data are hosted entirely within the EU (France and Sweden). No data leaves European jurisdiction. The platform complies with GDPR requirements and uses exclusively European AI providers for any language model integration.
The Observatory is designed as a public good. The PLI Index, sector trajectories, organisation profiles and headline findings are freely accessible without registration. Anyone can explore how institutional language is changing across European countries and sectors.
For researchers, the Observatory provides open access to aggregated PRISM™ scores and sector benchmarks. Institutional partners can contribute validated datasets and specialist corpora to strengthen the baseline and expand language coverage. Academic enquiries are welcome at research@copymama.it.
For organisations looking to understand their own linguistic trajectory in depth, CopyMama's Studio provides diagnostic engagements powered by the full PRISM™ pipeline. These extend beyond the public Observatory into VOICE™ governance, where diagnostic evidence translates into actionable editorial standards.
Academic partnerships, dataset contributions, methodology validation and publication collaboration.
PRISM™ diagnostics, VOICE™ governance, verbal identity consulting and institutional training.