Back
EU AI Act: Germany, France and Italy Reach Agreement on AI Regulation in Europe
webReports on a key political development in EU AI regulation: France, Germany, and Italy reaching a joint position on foundation model governance, which influenced the final EU AI Act negotiations and shaped how general-purpose AI would be regulated in Europe.
Metadata
Importance: 42/100news articlenews
Summary
France, Germany, and Italy agreed on a joint approach to AI regulation, supporting voluntary but binding commitments for all AI providers rather than only large ones. The three countries pushed for regulating AI applications rather than the technology itself, with a future European authority to monitor compliance. This agreement helped unblock stalled EU AI Act negotiations over foundation models.
Key Points
- •France, Germany, and Italy agreed that AI conduct rules should be binding for all providers—not just large ones—to avoid undermining trust in smaller European firms.
- •The three governments favored regulating AI applications rather than the underlying technology, to remain competitive globally.
- •Initially no sanctions would be imposed; a sanctions system could be introduced later if violations are identified.
- •A future European authority would be established to monitor compliance with AI standards.
- •The agreement came amid a deadlock between MEPs and member states over how to regulate foundation/general-purpose AI models.
Cached Content Preview
HTTP 200Fetched Apr 14, 20265 KB
By  Reuters
Published on 19/11/2023 - 13:21 GMT+1 • Updated
20/11/2023 - 11:58 GMT+1
Share
Comments
Share
Facebook
Twitter
Flipboard
Send
Reddit
Linkedin
Messenger
Telegram
VK
Bluesky
Threads
Whatsapp
Disagreements from key EU member states, including France, on how artificial intelligence (AI) should be regulated threatened to derail EU legislation.
France, Germany, and Italy have reached an agreement on how artificial intelligence (AI) should be regulated, according to a joint paper seen by Reuters, which is expected to accelerate negotiations at the European level.
ADVERTISEMENT
ADVERTISEMENT
The three governments support commitments that are voluntary, but binding on small and large AI providers in the European Union that sign up to them.
The European Commission, the European Parliament, and the EU Council are negotiating how the bloc should position itself.
Related
The EU's AI Act: A guide to understanding the ambitious plans to regulate artificial intelligence
In June, the European Parliament presented its AI Act designed to contain the risks of AI applications and avoid discriminatory effects, while harnessing the innovative power of AI.
On Friday, MEPs walked out of a meeting with member state representatives after reaching a deadlock over the proposed approach to foundation models, according to Eurativ .
France, Germany, and Italy were among the larger member states pushing against regulation, a move which threatened to derail efforts to get the legislation passed in this session of the European Parliament altogether.
AI rules should be binding for everyone
During the discussions in June, the European Parliament proposed that the code of conduct should initially only be binding for major AI providers, which are primarily from the United States.
The three EU governments have said this apparent competitive advantage for smaller European providers could have the drawback of reducing trust in them, resulting in fewer customers.
The rules of conduct and transparency should therefore be binding for everyone, they said.
Initially, no sanctions should be imposed, according to the paper.
If violations of the code of conduct are identified after a certain period of time, however, a system of sanctions could be set up.
Related
EU AI Act nearing agreement despite three key roadblocks – co-rapporteur
In the future, a European authority would monitor compliance with the standards, the paper said.
Germany's Economy Ministry, which is in charge of the topic together with the Ministry of Digital Affairs, said laws and state control should
... (truncated, 5 KB total)Resource ID:
2d32a52d746b63e3 | Stable ID: sid_69sIVxN7Fx