Which will just put pressure to improve the testing situation, not necessarily change stance on AI written patches. In fact AI is quite good at catching issues and root causing them back to recent changes so they could automate that journey a bit better as well to fill in any gaps not caught at review time.