Authorization

Like maybe I want, you know, my bot to be able to perform this action, but only on this subset of resources, you know, like as soon as you get, you can do that well in an AWS IAM, but like most other systems you can't.

So I think part of this is just, we're going to have more permissions for these than we want, at least in the medium term until we like redo AuthZ everywhere.

And so then it's like, okay, well, how do I control it? That would be like, you know, the review, that would be things like proxy where it just drops it conditionally, but that's also a lot of stuff to implement.

Yes, it's a lot to implement.

I would, um-- But if you haven't read the IETF draft on AI agent authentication and authorization, I would highly recommend reading that one because it just tries to cover everything at once.

That's a really good starting point.

And then you should definitely look at OpenID's AuthZ work and Shared Signals Framework work for continuous access evaluation protocol.

I'm throwing a lot of words here.

But essentially, it's OpenID and Shared Signals Framework group and IETF work.