I spent a lot of time thinking about this subject and it boils down to: I can try to do this in such a way that creates the best possible outcome or I can wait for someone else to do it in such a way that makes it worse for everyone.
Take something like https://writesonic.com/ as an example. It is backed by Y Combinator and claims to have 5M+ users using it to write articles. They do exactly what you described – they use your topic to generate an article cramming unreferenced/unattributed facts into whatever they claim to be "SEO optimized" article structure. In fact, one of the reasons for developing AIMD is because I saw a substantial uptick in copywriter submissions of what is very clearly Writesonic generated content. They don't care about factual accuracy or attribution.
In contrast, what I am building towards is a fact checked, properly referenced way of writing articles. Every borrowed fact, citation, quote, etc is properly attributed. You could obviously choose to remove those references, but part of what I want to achieve is educating that keeping those references actually increases the legitimacy of the content.
One way or another, a variation of this is going to dominate the vast majority of content writing. The best possible outcome is that someone creates tools that reward good behavior/good actors.