My point is that there probably is no way in principle to distinguish between a human user utilizing automation on their own behalf in good faith (e.g. RSS readers) and bad faith automations.