Via blogs.microsoft.com
The research found formal AI governance remains uncommon, with only 13% of companies worldwide reporting governance frameworks.
The EU’s Artificial Intelligence Act is increasingly shaping AI governance globally, with almost half of companies citing the law in governance disclosures located outside the European Union, according to a Thomson Reuters Foundation analysis.
The research examined 2,973 companies using more than 100,000 governance data points and found that 47% of organizations referencing the AI Act are headquartered outside the bloc. This suggests the legislation is already becoming an international benchmark before its most stringent rules take effect in August 2026.
The law applies not only to EU-based companies but also to organizations whose AI systems are used within the EU or affect European users.
The report found adoption remains relatively limited overall, with only 13% of companies maintaining formal AI governance frameworks.
Among non-EU companies, US firms accounted for the largest share of AI Act references, particularly in the technology sector, while major AI developers including Microsoft, Google, OpenAI and xAI have voluntarily adopted elements of the EU’s AI Code of Practice.
Despite improvements in governance, major compliance gaps persist ahead of stricter enforcement.
Only 12.4% of companies require human review of AI-generated decisions, and fewer than 25% evaluate AI’s potential impact on employee rights. Those requirements will become legally mandatory for high-risk AI systems in sectors including hiring, credit and healthcare once the Act is fully implemented.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our