Then that training will get called into evidence (like Google’s prior training re: anti competitive training) to prove they were being intentionally anti-competitive, and rinse repeat.
I’ve worked for one BigTech company in my career and there were a list of banned words we couldn’t say in writing. The one word I remember we couldn’t say was “moat”.
I received training like this over 20 years ago as a brand new engineer at my first job. The training was literally about how to avoid writing emails that could create legal liability due to careless language. Emails that came out during the Ford Pinto lawsuits were used as examples. It was all carefully worded to be about not being misperceived in court.
That job was at an avionics company working on safety critical systems. They paid tons of lip service to always placing safety first -- and from what I personally witnessed, at least at that time, the concern was genuine. There was a culture of taking the responsibility seriously, at least at the engineering level I interacted with.
Even acting in good faith though, things happen. Planes crash (usually due to pilot error), and when they do everyone gets sued, and when that happens careless language represents a risk for a company, even if they did everything right.
Having moved on to consumer tech, I haven't seen similar cultures of doing the "right thing". That could be the modern world, my own cynicism, or just the differences inherent to industries where lives aren't explicitly on the line. Regardless, it's not at all hard to imagine that employees can be taught to self censor in ways that won't themselves create more liability.