Reviewing my short post last night about the NYT vs. OpenAI data retention, maybe “wild” overreach was an unnecessary adjective. I also hadn’t seen OpenAI’s response:

As part of their baseless lawsuit, they’ve recently asked the court to force us to retain all user content indefinitely going forward, based on speculation that they might find something that supports their case.

There are very real fair-use questions for training, but I’m not sure they can be resolved by this lawsuit, and probably not without updating copyright law.