Emerging EU AI Act: Pioneering a Global Standard for Artificial Intelligence Regulation

In April 2021, the European Commission proposed a comprehensive legal framework on artificial intelligence (AI). The primary objective of this legislation is to regulate the use and development of AI in the European Union (EU).

Known officially as the European Union Artificial Intelligence Act, or EU AI Act, it is expected to pass later this year. According to estimates, the Act will likely be enforced two years after approval, possibly at the end of 2025 or the beginning of 2026.

The significance of the EU AI Act lies not just in its implementation within the EU, but it also sets the tone for global standards around AI regulation. The principles and guidelines that it introduces may inform and influence the AI policies of other countries as they come to grips with the complexity of governing AI innovations.

In brief, under the Act, AI systems are divided into several categories depending on their potential to harm individuals or society. For each category of AI, the Act lays down certain restrictions and requirements.

Unquestionable AI applications that manipulate human behaviour or have a potential negative impact on people’s safety, livelihoods and rights are prohibited outright. This includes AI used in deceptive and subliminal marketing techniques, and social scoring systems by governments.

High-risk AI applications — those whose risks can be managed through strict compliance with specific mandatory requirements are also addressed in the Act. These requirements cover data governance, transparency, technical documentation, human oversight, security and accuracy. AI in critical infrastructures, educational systems, employment and essential public services fall under this category.

The Act also creates protective measures such as the establishment of national supervisory authorities, a European Artificial Intelligence Board, and the development of codes of conduct. In this way, the Act is attentive both to the innovation and implementation of AI technology, and also its potential risks and harm.

With its passage, the EU AI Act will not only shape the future of AI in Europe but is poised to significantly influence international law and policy regarding the governance of AI technologies. Such global impact points to the Act being an emerging global standard for AI.