You ban open weight models and the problem is solved

Users of the model could be legally required to take part in the tracking.

Just the same? If you publish a model that doesn't follow these rules nobody in the EU could use that model in their business. You could publish unlicensed source code as well and nobody could really use it for anything business related either.

> If you publish a model that doesn't follow these rules

The rules require tracking outputs, which open-weight models cannot do. So I'm wondering if open-weight models have separate rules or this effectively bans releasing such models.

Of course they can. Lets assume you are using such a model in your product, this now makes tracking its output your responsibility. It is really no different from the way you would use an open source library.

That is confusing since closed-weight models (the models, not the applications using the models) also can't track outputs. It would be weird if the rules applied to the model and not the application because then it literally only applies to open-weight models since the closed-weight models are, by definition, never released to the public.

Trying to understand the rules but it doesn't seem to make a clear distinction between these things. I assume that they are intending the applications that use the models, not the models.

Why would it be the creators’ responsibility? It should be the one running the model.

The service provider (or whoever was _running_ the thing) would be liable, I'd assume, not the model creator.