Is claw-generated code copyright-able?

Given the US Supreme Court's decision to leave a lower court ruling in place, it seems that the Copyright Office's guidelines (humans must be involved or it's not eligible for copyright) are the official law of the land. That means that code (or other "works") generated purely by prompting an LLM (or made autonomously by one) are not eligible for copyright.

Does that mean that all code generated by the OpenClaw agents, can't be "open source"? (because it's not eligible for copyright at all) What about reports/code/etc generated autonomously by LLM agents? Can any company claim code as an asset in an acquisition if the code itself isn't own-able (because it's public domain according to the copyright office)?

I confess to being surprised at the repeated "coding agents wrote 90% of our code last quarter" statements from CEOs...those sound like court-admissible statements that they don't own copyright on their own code.