Interesting point of view. Mind elaborating on why you think so?
Google make money on Ads.
That is their #1, primary and main source of income.
They need to moderate enough to stay "reputable" and stay out of legal trouble.
Everything else honestly, is free game to them.
People may think Google hates their websites, seo etc etc but honestly, Google most likely really doesn't even know of your existence & have no real reason to target you either.
By making us all chase these constant SEO "algos" updates makes us feel like the game is still fresh and active & for people to contiune exploring the "wonders of SEO". /gags
To them, it's probably the aftermath results of them tweaking and polishing up their ads to optimize more revenue.
That's the first point to why even if Google has a failsafe AI detector, they won't use it at 100% to ban or blacklist websites.
The second point is
I can write an article that sounds and feel like an AI
An AI could be instructed to write and article as closely to human as possible.
With those 2 statements, there are infinite possibilities.
For example
I can feed and train my model with constant high school students showing it between students who write articles that are A, B and even C. then instruct an AI to write an article based on it's grades.
Would AI be able to tell, between an article that is written by a student who is just struggling with the English Language, or written by an AI attempting to mimic that?
Which then ties back to #1
Would Google really risk blacklisting websites that are potential revenue just because an article was written by AI?
I highly doubt it.
That's most likely why Google is going the approach of "Helpful Content" vs "Generated content for the purpose of ranking".