I should declare a bias before I start. I grew up in regional Austria surrounded by agricultural co-operatives. The Lagerhaus sold the seed, the fertiliser and the diesel. The dairy co-op collected the milk. I’ve known them from the farm gate and, briefly, from inside the farmers’ league that runs alongside them. I’ve seen what they do well when their members actually run them, and I’ve seen what happens when governance gets comfortable and they start behaving like the corporations they were founded to keep at bay. They remain, perhaps sentimentally, a favourite of mine.

There are two dominant ways to think about who should govern artificial intelligence.

The market frame says: the data, the models, and the compute are private goods, owned by the organisations that assembled them, exchanged through contract and property rights. Governance follows ownership. OpenAI owns GPT-4. Google owns Gemini. These are products. Markets will sort it out.

The state frame says: AI is too consequential to leave to markets, so regulators must impose rules from above. The EU’s AI Act is the most developed version of this: binding requirements, conformity assessments, registration databases, penalties for violations.