The thing is that, according to the DailyDot article that's been posted around here, the account that lead to The Atlantic being banned was posting hundreds of links to The Atlantic and their subsidiaries. What's non-trivial about detecting that?
It becomes non trivial to detect when they start spamming using 100s of aged accounts run through proxies. You can detect that lots of people are submitting stories to the atlantic, but it looks organic.