Because regulating AI isn’t just about (product) compliance. It’s about democracy, human rights, and global trust.
Link to the original report
 

Here’s the distinction:

The AI Act (EU law)
• A rulebook for the European market.
• Detailed obligations for developers and deployers.
• Bans certain harmful AI uses.
• Enforced with audits, conformity checks, and fines.

The AI Convention (Council of Europe treaty)
• A global baseline of principles: dignity, equality, non-discrimination, democracy, rule of law.
• Binds governments as well as companies.
• Requires risk and impact assessments, remedies for harms, and even permits/bans/moratoria on AI uses that undermine rights.
• Open to any country in the world (the US, UK, Israel and others have already signed alongside EU states).

In short: the AI Act keeps Europe safe, while the AI Convention sets the standard for the world

It’s like this:

The AI Convention is the traffic code - universal principles keeping everyone on the same road.
The AI Act is the seatbelt - practical, technical, enforceable.

The EU isn’t duplicating effort. It’s leading on two fronts: rules at home, principles abroad.
 

Question: Do you think this two-layered model - national/regional rules + global principles - is the future of AI governance? Or will national interests always block true global standards?