I saw that on day one.
Perhaps on the frontend some functions can be muted combined with the abuse mitigation that comes standard with that service. Not sure how that would gel though when scaled up. Suppose an unauthorized operation could send a command to the service and shut it down for that account, but I dunno.