- EU establishes new AI Office to oversee implementation of upcoming AI Act.
- The office comprises five units focused on regulation, safety, innovation, social good, and policy coordination.
- The office is already working on guidelines and codes of practice for AI systems.
- The EU aims to balance AI innovation with responsible oversight.
Brussels is bracing for the dawn of a new era in artificial intelligence governance. With the EU AI Act slated to take effect this summer, the bloc has unveiled the structure of its newly established AI Office, a body tasked with fostering the European AI ecosystem while also keeping a watchful eye on potential risks.
The AI Office, launching on June 16th, embodies Europe’s dual ambition: to be a global leader in AI innovation and a standard-bearer for responsible AI use. The office’s structure, comprising five distinct units, reflects this complex mandate.
Five units, one mission
- Regulation and compliance: This unit will be the enforcement arm of the AI Office, ensuring harmonized application of the AI Act across EU member states. It will investigate potential violations and administer sanctions, working in tandem with national-level regulatory bodies.
- AI safety: With the recent explosion of generative AI technologies like ChatGPT, this unit will focus on identifying and mitigating the systemic risks posed by powerful general-purpose AI models. It will conduct testing and evaluations, particularly for models trained on vast amounts of data, which the AI Act designates as having “systemic risk.”
- Excellence in AI and robotics: This unit will champion AI research and development, coordinating with initiatives like “GenAI4EU,” which aims to boost Europe’s generative AI capabilities.
- AI for social good: This unit will spearhead international collaborations on AI projects that have the potential to benefit society, such as weather modeling, cancer diagnosis, and cultural preservation.
- AI innovation and policy coordination: Tasked with ensuring the execution of the EU’s broader AI strategy, this unit will monitor trends, stimulate AI adoption, and foster a vibrant innovation ecosystem.
Balancing innovation and oversight
The structure of the AI Office, with three of its five units dedicated to fostering AI development, sends a clear message: Europe is committed to both innovation and responsible oversight. The bloc believes that trust is essential for widespread AI adoption, and its approach aims to strike a balance between fostering cutting-edge technology and safeguarding against potential harms.
While several key positions within the AI Office have been filled, including the head of the office itself, the crucial role of the AI Safety unit chief remains vacant. The office is also in the process of hiring additional staff, with plans to reach a headcount of over 140 in the coming years.
The AI Office’s work is already underway, with guidelines on AI system definitions and prohibitions in the works. The office is also preparing codes of practice for general-purpose AI models, which are expected to serve as an interim guide while the AI Act is phased in.