This appears to me a great public and private tech rabbit hole. I entered with the questions, "and then what AIs will the surveillors train?" and "what's been published on the technologies this defeats?" The first link in the article is https://www.popularmechanics.com/technology/security/a38147393/ai-hack-brain... which is an article on how current AIs are engaging in automated social engineering and compromising the human biosystem for profit.