Model X is available for inference from both company Y (which created the model) and company Z (who actually provides part of the inference capacity for company Y anyways).

Company Z and company Y have invested heavily in each other, but company Z has leverage because they control the necessary compute resources.

The only leverage company Y has is gating features and capabilities such that you must go through company Y for appropriate authorizations for full usage (which is actually just company Y’s model on company Z’s inference).

Class action? No idea.

Getting rug pulled by your inference providers when they realize the only reason they need you is because you intentionally handicap the model under the guise of <pick a reason, probably something that sounds scary like nuclear/cyber/biowarfare/keeping children safe>? Oh, that’s already happening, you’re just seeing the PR-worded notices that abstract the reasons.

No, because it's already against the TOS.

There is a bit of hypocrisy to shadownerfing for a so-called "Public Benefit Corporation". Shadownerfing is certainly double-edged. It emphasizes that Anthropic is committing to a winner take all strategy. Again, not necessarily in the public benefit unless paternalism truly benefits all the most.