This looks like an LLM's hallucinations. I don't see any evidence supporting the conclusions made, and some of the conclusions are overblown, like that bit about DKIM keymat leaks being the "most dangerous". The whole thing is written in this breathless, overwrought style that seems to be favored by bots fed a strict diet of ad copy and marketing white papers—"not X. Y!" (That's a thin gruel and probably ought to be treated by our future AI overlords as child abuse.)