There are fundamental issues with the way AI is being promoted and used by companies that cause significant concerns. Using AI as a tool makes sense, but because AI does not have a fundamental ability to test itself and the results it generates, you should not automatically trust what comes from AI. Things like "AI hallucinations", where AI gives things like legal cases that do not exist as an argument for a legal brief are documented. The structure of software written by AI may also defy the ability to fix problems in the code generated. If the code generated by AI can't have bugs fixed because the code isn't well designed/structured, that's a BIG problem.
And then, there is the "over-promise" of AI...yea, AI can be used to eliminate human jobs, before the actual quality of what AI is doing has been verified. This would be like taking a group of high school graduates, giving them jobs, but then, not having experienced people monitoring what those recent high school graduates were doing, or correcting "bad behaviors".
Again, AI has potential, but is too new to be trusted at this point to be a full replacement for human employees. As a tool for employees, hey, it's all good, but those employees should also do a verification that what the AI came up with is accurate.