The EU AI Act: Balancing Innovation and Ethics in Artificial Intelligence Regulation

The EU AI Act is a proposed legislative framework aimed at regulating the development and use of artificial intelligence technology within the European Union.

The act aims to ensure the ethical and responsible deployment of AI systems while striking a balance between encouraging innovation and protecting individuals’ rights.

The draft law would limit how companies can use AI to protect privacy, and require that they implement the technology in a safe and transparent manner. It also establishes a risk-based ranking system that would force companies to disclose the potential risks of their AI products to users, and requires them to obtain prior approval for certain AI uses.

AI with minimal risk, such as spam filters and video games, would be exempt from these transparency requirements.

However, the draft law has several serious flaws. It does not adequately address the role of the public in the process of designing, developing, and deploying AI systems.

Furthermore, it does not adequately address how the law will be enforced or how it will impact the economy and employment.

In particular, the law’s attempt to ban social scoring—a practice that evaluates a person’s characteristics and behavior through various data sources—will have significant repercussions on people’s ability to find jobs, access loans, and maintain housing.

MEPs also expanded the categories of “high-risk” AI systems to include those that harm a person’s health and safety, their fundamental rights or freedoms, democracy, or the environment.

They included recommender systems used by social media platforms, and AI used to influence voters in political campaigns.

They also included a new category of “generative foundation models” —AI that generates content (e.g. ChatGPT) —which are already subject to strict regulations in Europe.

The compromise text also contains a number of exceptions that will allow Member States to apply national rules that go beyond the EU provisions, and it gives European law enforcement institutions wide latitude in determining how and whether to use AI in their work.

As a result, it will be difficult for the EU to implement this proposal in full without significant resistance from Member States and from other stakeholders.

Nonetheless, this is only the first step in the long process of building the EU’s comprehensive legislative framework on AI. It will take years to complete, and it is likely that foreign governments will push for similar legislation in order to shape international standards.

The EU is also promoting efforts to encourage cooperation between private and public bodies in the creation of AI. These proposals reflect a clear commitment by the EU to promote cooperation on AI and global governance in general.

This will be a critical test of whether the EU can become an international leader in addressing the challenges and opportunities presented by the rapid development of AI.

If not, the EU will be left behind by the rest of the world. Frances Burwell is a distinguished fellow at the Atlantic Council’s Europe Center and senior director at McLarty Associates.