OpenAI published new Signals data on June 30, 2026. The ChatGPT usage data reveals a massive shift in global artificial intelligence adoption. More than half of active ChatGPT users on individual consumer plans now predominantly use a language other than English. This dataset covers Free, Go, Plus, and Pro plans. It explicitly excludes Enterprise, education, and Codex usage.
Increasing ChatGPT Usage in New Regions & Languages
ChatGPT adoption has grown sharply across every continent since July 2023. However, Africa and Asia experienced the fastest relative growth. Consequently, lower-Human Development Index (HDI) countries saw the largest user base increases. OpenAI credits this rapid adoption to its continued provision of low-cost access options.
English usage dropped to around 40% of the overall user share. Meanwhile, Spanish, Portuguese, and Arabic lead the expanding non-English majority. Furthermore, smaller languages display explosive growth. Search Engine Journal confirmed that among languages with over one million users, Uzbek, Kazakh, and Burmese grew the fastest.
Deepening Engagement & Demographic Shifts
Users interact with the AI much more over time. Six months after signing up, users send 50% more messages per day. They also double the number of distinct tasks they try across 53 different classification categories. The data sample analyzes 0.1% of accounts created between October 15, 2025, and May 1, 2026. It tracks activity up through May 31, 2026. The analysis excludes minors, banned accounts, and completely inactive users.
Additionally, demographic data highlights a significant gender skew. People with typically feminine names now represent the majority of global usage. Countries like Brazil, Colombia, Poland, and Namibia lead this female usage skew. Conversely, Pakistan, Bangladesh, Angola, the Democratic Republic of Congo, and Mali show heavy male usage concentration. OpenAI estimates these figures using name-to-gender crosswalks, as the company does not collect direct user gender information. Unclassifiable names are omitted from the data.
Technical Implications for AI Teams
This linguistic shift fundamentally changes priorities for AI builders. AI teams must now reassess their evaluation standards. Generic English-centric benchmarks systematically undercount errors and blind spots. Therefore, developers urgently need robust multilingual test suites, language-aware tokenization, and per-language telemetry.
Moreover, the rapid growth in lower-HDI regions demands direct infrastructure changes. Engineers must prioritize lightweight, latency-tolerant models. They also need to build mobile-first user experiences optimized for low-bandwidth environments. Moving forward, the industry should watch for localized user interfaces, targeted pricing tier adjustments, and language-specific safety metrics from major AI competitors. Third-party datasets for fast-growing languages will also become critical for downstream fine-tuning.
