I mean, the problem isn't that they'll look like hypocrites...

The problem is that the bullshit they spout can be used against them in court.

For instance, some of these companies are on trial for using copyrighted works for training. One could claim that the training of a model is fair use. One of the criteria for establishing fair use, is that it doesn't damage the market for the original work.

But their email claim they think LLMs can effectively damage the creators of the content they use, and they publicly repeated over and over how their AIs will make everyone unemployed.

Now they could say that they bullshitted the public, but then MSFT is a public corp, so that might be considered market manipulation? And indeed Microsoft has walked that back earlier than most others.