ACT, Artificial Intelligence Act, AI Law

Understanding the European Union’s AI Law

In a significant move, European Union lawmakers have recently given their final approval to the world’s first comprehensive artificial intelligence (AI) law, known as the Artificial Intelligence Act. This landmark legislation, which is set to come into effect later this year, aims to provide a framework for regulating AI technology within the 27-nation bloc. Let’s take a closer look at what this new law entails and its implications.

What is the AI Act?

The AI Act takes a human-centric approach to AI regulation, emphasizing the importance of ensuring that humans remain in control of the technology. It adopts a risk-based approach, meaning that the level of scrutiny applied to AI systems depends on their potential risk factors. While low-risk AI applications, such as content recommendation systems, are subject to voluntary requirements, high-risk uses, like medical devices or critical infrastructure, face stricter regulations.

Banning certain AI uses

One of the notable aspects of the AI Act is its prohibition of certain AI applications deemed to pose unacceptable risks to individuals or society. For instance, the law explicitly bans the use of social scoring systems and predictive policing algorithms. Additionally, the use of AI-powered biometric identification systems for public surveillance is restricted except in cases of serious crimes. These measures aim to safeguard individuals’ rights and prevent the misuse of AI technology.

Regulation of generative AI

Recognizing the emergence of general-purpose AI models like OpenAI’s ChatGPT, the AI Act includes provisions specific to generative AI. Developers of such models must disclose the data used for training, comply with copyright laws, and label AI-generated deepfakes to indicate artificial manipulation. By imposing transparency and accountability measures, the law seeks to address concerns related to the proliferation of deceptive or harmful content generated by AI systems.

Oversight and mitigation of risks

In addition to specific prohibitions and requirements, the AI Act imposes additional scrutiny on large AI models that pose systemic risks. Companies developing or deploying these models must assess and mitigate potential risks, report incidents, implement cybersecurity measures, and disclose energy usage. By holding companies accountable for the potential harms associated with AI technology, the law aims to promote responsible innovation and ensure the safety and well-being of users.

Global impact

The EU’s adoption of comprehensive AI regulations sets a precedent for other jurisdictions grappling with similar challenges. While the United States and China have also taken steps to address AI governance, the EU’s approach is notable for its emphasis on protecting human rights and mitigating risks associated with AI technology. By establishing clear rules and oversight mechanisms, the EU seeks to influence global standards for AI development and deployment.

Implementation and enforcement

The AI Act is expected to become law by mid-2024, with provisions phased in over time. Each EU member state will establish its own AI watchdog for enforcement, while a centralized AI Office in Brussels will oversee compliance with regulations for general-purpose AI systems. Violations of the AI Act could result in significant fines, underscoring the EU’s commitment to ensuring adherence to AI regulations and protecting the interests of individuals and society.

Looking ahead

While the AI Act represents a significant milestone in AI regulation, it is not the final word on the subject. Lawmakers anticipate further legislation, particularly in areas such as AI in the workplace, following upcoming elections. The EU remains committed to fostering innovation while ensuring responsible and ethical AI development. By continuing to engage stakeholders and adapt regulatory frameworks to evolving technological advancements, the EU aims to maintain its leadership role in shaping the future of AI governance.

If you would like to find out how AI works in SAP then read our recent blog articles. Let’s bring AI into your business today!

SEE HOW WE CAN DIGITALIZE YOUR BUSINESS

Let's meet on social media: