I'd love to have this as a norm, as long as the AI doesn't lie and impersonate and deceive humans in speech by saying things as though it were the human like "I ate breakfast" or "I was saying this to my team" because clearly the AI did not do those things. That's lying, deception and harmful to a culture of truth and transparency.

Instead the humans who promoted and allowed the LLM content to post behind their human identity, didn't bother to update the LLM language and either do a) mark the post as AI generated or b) properly update those pronouns so it isn't an AI speaking through the humans point of view.

Consider this exaggerated example: Would it be ok for you if in a zoom meeting with your team someone was lip syncing an AI speaking on their behalf, both impersonating voice tonality, the words chosen, and even pretending to voice the humans thoughts themselves? Of course you wouldn't. So now extend this to the words people write in articles like this one where the "I said" perspective was used many times supposedly by an AI.