EU AI Act: Thierry Breton on Regulation, Innovation in Japan

EU Finalizes Landmark AI Regulation After Council and Parliament Agreement

EU Commissioner Thierry Breton outlined the final stages of the revised AI Act in an interview with NHK World's Yoshioka Takuma. The legislation has secured agreement from the European Council representing all 27 member states and the European Parliament. Once formally adopted, the Act is projected to become final and enforceable within two or three months from the time of the interview.

This timeline reflects the standard EU legislative process following political agreement. The regulation introduces a structured framework for artificial intelligence across the single market, with particular attention to generative systems and transparency requirements.

Risk-Based Framework Targets Specific High-Risk Uses

The AI Act follows a risk-based model. Most AI applications face no additional constraints under the new rules. A smaller set of uses involving sensitive data training must meet transparency and documentation obligations. A narrow category of practices, such as social scoring systems, is prohibited outright.

Generative AI models receive dedicated obligations, including mandatory watermarking for AI-generated images, text, and video. These measures aim to improve traceability without imposing blanket restrictions on lower-risk tools. Breton emphasized that the approach seeks to balance risk reduction with continued technological development.

Generative AI Obligations and Industry Engagement

Breton confirmed that OpenAI chief executive Sam Altman has stated his company's willingness to comply with the Act's requirements. The obligations for generative systems focus on labeling and technical safeguards rather than pre-approval processes for most deployments.

An open letter signed by more than 150 European companies, including Siemens, Airbus, and Renault, raised concerns that the rules could affect Europe's competitive position. Breton countered that clear regulation provides visibility and predictability, drawing an analogy to mandatory safety features such as seatbelts and airbags in the automotive sector.

Japan's Distinct Regulatory Path and Upcoming G7 Summit

Japan has pursued a different strategy. Its AI Act, passed in 2025, emphasizes basic principles and non-binding guidelines rather than strict penalties. The Japanese government, through the Digital Agency and METI, continues to prioritize innovation while monitoring international developments.

Breton specifically welcomed Japanese companies to the EU market, noting that they would need to meet the same obligations as other participants but indicating that compliance should not present significant difficulties. Japan plans to host a G7 AI summit in late 2026 aimed at promoting interoperability among Japanese, EU, and US regulatory regimes.

Global Governance Efforts and Timeline for Labeling Rules

The G7 has advanced discussions toward a shared perspective on AI governance. Breton expressed readiness to share the EU's regulatory experience with partners including Japan and the United States. Separate from the AI Act itself, the EU's Code of Practice on marking and labeling AI-generated content is scheduled to apply from August 2, 2026.

Enforcement of related legislation, including the Digital Services Act, began in August, while the Digital Markets Act continues its phased rollout. These measures form part of a broader digital regulatory package affecting platforms and services operating in Europe.

Economic Context: Market Size and Semiconductor Demand

Breton highlighted that the European digital market is one-and-a-half times larger than the US market, underscoring the commercial importance of regulatory clarity for companies seeking access. The global semiconductor market is projected to increase by 90 percent in 2026, driven substantially by AI-related demand.

Japanese firms active in semiconductors, automotive electronics, and industrial automation will need to assess how the Act's transparency and watermarking requirements intersect with their product roadmaps. METI's ongoing work on Green Transformation and Society 5.0 initiatives provides a domestic policy backdrop against which international compliance costs will be evaluated.

By Kenji Tanaka, Staff Writer