EU AI Act introduces new global benchmark for regulation

The imminent enforcement of the European Union's Artificial Intelligence Act on 2nd February marks a significant step in the regulation of AI technologies, as emphasised by expert commentary from various sectors. The Act's initial focus is on prohibiting high-risk AI use-cases that could threaten core EU values. Such prohibitions include the use of AI for social scoring, emotion recognition, and facial recognition, which are now under strict surveillance.

Enza Iannopollo, Principal Analyst at Forrester, highlighted the importance of the milestone, stating that the Act aims to combat the negative impacts of certain AI technologies on fundamental rights. Although the enforcement begins now, the full scope of sanctions remains in preparation, indicating that significant fines will not be issued immediately. However, the lengthy compliance timeline suggests increased regulatory actions by mid-year, with potential fines reaching as high as 7% of global turnover for non-compliant entities.

Leslie Joseph, also a Principal Analyst at Forrester, noted the global implications of the EU AI Act, which sets a new benchmark for AI governance that companies worldwide must adhere to. This regulation also poses direct challenges to businesses in Asia, particularly those in India, China, and South Korea, that are engaging in AI exports to the EU. Joseph pointed out that Asian companies must align with the EU's stringent criteria on transparency and risk management, which is driving discussions around regulatory harmonisation in Asia to maintain international trade relations.

In contrast to the fragmented approach within Asia, where regulations often rely on voluntary frameworks, the EU AI Act provides a unified regulatory strategy. Some Asian nations, such as South Korea and Japan, are moving towards more comprehensive AI governance, with legislative proposals aligning with global norms and discussions on harmonising AI policies, respectively.

The impact of these regulations is particularly pronounced in the financial services sector. Diyan Bogdanov, Director of Engineering Intelligence at Payhawk, discussed Europe's role in defining the standards for AI applications in finance, especially those deemed high-risk like credit scoring. Bogdanov remarked that these regulations do not constrain innovation but rather establish a necessary framework for developing trustworthy and reliable AI systems, positioning Europe as a leader in the field.

Complementing these insights, Levent Ergin, Chief Strategist at Informatica, underscored the importance of robust data governance to meet regulatory demands and ensure AI delivers meaningful business outcomes. With the EU focusing on data quality and governance, businesses must work diligently to manage their data effectively and ensure it is AI-ready. The upcoming months will test organisations' capabilities to balance compliance with achieving tangible returns on their AI investments.

The EU AI Act represents a comprehensive approach to AI regulation, setting expectations for organisations worldwide and establishing a precedent for governance that other regions might follow. The unfolding of these initial steps will likely have a lasting influence on AI development and management practices both within the EU and globally.