Global leader in technology advocates for immediate development of international regulations on AI usage
In a recent development, U.S. President Donald Trump unveiled an AI strategy that calls for sweeping deregulation to ensure the United States stays ahead of China in AI development. However, the need for a global approach on regulating artificial intelligence (AI) has been emphasised, with Doreen Bogdan-Martin, the head of the International Telecommunications Union (ITU), highlighting the urgency of getting the right framework in place for AI regulation.
The European Union (EU), China, and the United States (US) have distinct global approaches to regulating artificial intelligence, reflecting different priorities, governance models, and levels of regulatory detail.
European Union (EU)
The EU leads with the first comprehensive AI legal framework, the EU AI Act, effective from 2024. This framework uses a risk-based approach, categorizing AI systems into unacceptable, high, limited, and low risk, with strict rules especially for high-risk AI. The EU's key regulatory goals include protecting fundamental rights, safety, transparency, and preventing manipulative AI.
The EU's regulatory framework is enforced by the establishment of a European Artificial Intelligence Board for coordinated enforcement among Member States. Binding conformity assessments are required for high-risk AI systems prior to market placement.
China
China emphasizes centralized control and social governance, blending regulation with strategic industrial policy closely linked to state security and political objectives. The country has strong government-led controls, with detailed technical standards, mandatory data and algorithmic audits, and state oversight to ensure alignment with national interests and ideology.
United States (US)
The US favours a decentralised, sector-specific regulatory environment prioritizing innovation and flexible governance. There is no overarching federal AI law; regulation is spread across agencies dealing with privacy, consumer protection, and national security, supported by voluntary guidelines and frameworks. The US plan includes more than 90 proposals, one of which is the promise to "remove red tape and onerous regulation" that could hinder private sector AI development.
A Global Dialogue on AI Regulation
Doreen Bogdan-Martin, the head of the ITU, emphasized the importance of global dialogue on AI regulation. The debate on AI regulation should involve different national approaches, including the EU, Chinese, and U.S. approaches.
The tripartite contrast reflects the EU's regulatory ambition to shape ethical AI on a global scale, China's state-driven control model, and the US market- and innovation-driven architecture. Concerns over the risks posed by AI include mass job losses, the spread of deepfakes and disinformation, and society's fabric fraying.
A consistent theme among existing AI strategies is a focus on innovation, capacity building, and infrastructure investments. Doreen Bogdan-Martin did not comment on her concerns about an approach that urges less, not more, regulation of AI technologies, stating that she is still trying to digest the U.S. plan.
[Sources: 1, 2, 3, 4, 5; supplemented with general geopolitical context knowledge.]
Technology and artificial intelligence (AI) are at the center of global discussions regarding regulations to ensure ethical and safe development. The European Union (EU) is leading with the first comprehensive AI legal framework, the EU AI Act, which categorizes AI systems by risk and prioritizes protecting fundamental rights, safety, transparency, and preventing manipulative AI. In contrast, China focuses on centralized control and social governance, blending regulation with strategic industrial policy, while the US favors a decentralized, innovation-driven regulatory environment with less oversight. A global dialogue on AI regulation is deemed necessary to address concerns over risks posed by AI, such as job losses, deepfakes, and disinformation, and to shape ethical AI on a global scale.
