State of AI in Norway 2026

This report describes the state of artificial intelligence in Norway in mid-2026 across adoption, workforce, talent, domestic model production, compute, public investment, policy, public sentiment, responsible AI and safety, and education. It uses the public data of the 2026 Stanford AI Index and adds additional data collected between 2025–2026 that the AI Index report did not cover. On adoption, Norway records AI diffusion of 46.4% (third globally), 67% of workers using AI at work, and 61% trust in workplace AI. On production measured at the global frontier the original report and the data register a single “notable” model in contrast to an inventory of models led by Norway-based organisations identifies 21 model families, concentrated in Norwegian-language, speech, and applied scientific systems, which we identified for this report. Over 2025–2026 the state funded six national AI research centres, opened a national centre for AI security and safety, upgraded the national supercomputer, and saw large AI data-centre projects announced; a national AI law implementing the EU AI Act through the EEA Agreement is expected to enter into force in the second half of 2026. Norway does not train general-purpose foundation models, and its public AI investment of $716 million is below that of Sweden and Finland and close to Denmark. We also report the Centre’s own adversarial testing of Norwegian and frontier AI models: safety under sustained pressure, and in particular going along with confident false claims in medicine, remains an unsolved, industry-wide risk that anyone deploying AI in Norway is exposed to.