No one can tell Russian bots from average trolls on Twitter, and that’s a very bad thing

Twitter’s decision to simply delete the tweets of accounts it says are part of a Russian-linked propaganda bot army has proven to be really short-sighted because, well, it turns out that those tweets are really similar to your average Twitter troll. And telling the difference is a crucial, if difficult, task that the average user needs to be able to do in order to operate more safely on the platform.

Look no further than the @TEN_GOP account, assumed by most Twitter users to have been an account associated with the Tennessee GOP. It was cited numerous times by major mainstream news outlets throughout the 2016 presidential campaign and, yet, was ultimately determined to be one of the many offending accounts. 

Or take into account that, so far, Twitter has notified 1.4 million users that they’ve somehow engaged with accounts run by the Russian-linked Internet Research Agency (IRA). That’s a hell of a lot of people, particularly when we’re talking about influencing a closely fought presidential election. 

It’s good, then, that NBC News has published a database of thousands of tweets and accounts that were flagged by Twitter as being part of the Russian effort, because it shows you how hard it is to differentiate the bots from the actual Twitter users. (We collected some examples to give you an idea.)

From @jenn_abrams, one of the most notorious of these bot accounts (this tweet was retweeted over 500 times):

Liberals keep on saying that Islam is peaceful while Muslims firebomb bus in Paris https://t.co/VQ2MHPxikD 

Another @jenn_abrams gem that was retweeted 485 times and had more than 200 favorites:

Group of immigrants attacked a couple in Italy Europe, what’s wrong with you? #refugeeswelcome  https://t.co/0D82UudF0X

And here’s one from @TEN_GOP that was retweeted almost 1,300 times:

Reminder: Donald Trump was never called a racist in 68 years until he opposed Hillary Clinton for president! https://t.co/Xc1CqQFr05

Another account flagged by NBC, @traceyhappymom, showed how these accounts incorporated trends and topics beyond politics to help them blend in: 

Dr. Fin, Medicine Woman #FishTV @midnight

Okay, that’s clever. But it doesn’t change the fact it comes from a Russian bot farm. And this is why Twitter’s lack of transparency is so disturbing: these tweets are often almost indistinguishable from your regular, run-of-the-mill human trolls. (Not even the poor grammar is that different from the postings of a group notorious for “your/you’re” mistakes.)

Without being more transparent about their process and findings, Twitter is damning us to make the same mistakes over and over. If we can’t learn to distinguish the Russian bots from the real — and if Twitter won’t share that information — they’re  driving users away from a platform that, at best, is too secretive and, at worst, is guilty of egregiously encouraging those who want to use that platform to undermine democracy.