Twitter troll bots are not new. Reports of them have been surfacing in the tech media world for years now. They are bits of computer code which comb through tweets looking for keywords, hashtags, and phrases. They can be programmed to follow certain Twitter handles, as well. When they see one they have been programmed to spot, they respond with something nasty designed to intimidate, frighten, etc. It is important to realize these are bots, and not people, when something like this happens to you.

There are real people using Twitter. You can tell, very often, because they respond as humans do when queried. For example, recently I had a nice, brief Twitter conversation with someone whose profile said he is a minister in North Carolina (profiles can be fake, but I took him at his profile, so to speak) about a post of his. I thought the title of his post (which he said an editor wrote, not him) was misleading. I suggested, via a reply tweet, that he might ask the editor to reconsider the headline. He liked my reply (clicked on the heart icon) and we both moved on.

With a bot, though, you will never get a sensible reply to your question or query. All you ever get is insults or nonsense. Now, I know that is all you ever get from some people, but you get my point.

Onward.

Yesterday and today I have been having a “conversation” with a crude example of a Twitter troll bot. Just for fun I decided to poke the troll and see what it came back with. I am sharing part of the conversation here as an example of this kind of digital interaction along with some tips on how to spot a crude bot (some of the sophisticated ones are harder to spot btw). I also want to pass along a few tips on how to respond and how to shut one down.

This interaction started when I tweeted a link to a story about what Hillary Clinton said yesterday about Putin. I think the words in bold appearing in the same tweet was enough to attract the bot’s attention.

The bot’s (its name is Mona M) response to the tweet was classical troll language: @cnni @thejossgroup @realdonaldtrump Hillary’s own words” what difference does it make now”! Move on ~ Bernie DID when YOU pulled the rug…

Notice in the response the bot has added Donald Trump’s handle. I am not sure why, perhaps this was added as a tracking or counting mechanism.

Almost immediately after this tweet came another response tweet, one which mentions something not related to my initial tweet: @thejossgroup NO conflict of interest just another SMART decision to have HIS children sit in on a technology Apple/ Facebook meeting. SMART

I had not mentioned in that tweet the fact Trump had his kids sit in on his meeting with Silicon Valley execs. But maybe the troll bot just added that for good measure.

So, now we start to have fun (or at least I did) with a series of my reply tweets and Mona M’s replies. At one point I tweeted Mona is a bot, and the reply made me laugh out loud, “What is your language?” I got a couple more generic slam tweets as part of the exchange, such as @Reuters @thejossgroup Abedin should REVIEW a few things she’s been involved with…… and @thejossgroup Elizabeth has always been DEEPLY troubled. Mental health is a priority for President elect Trump. Kanye knows all about it.

Some of the replies to me poking the troll bot seemed to be sentences strung together at random (see paragraph above for examples). I am also uncertain of why all caps were used on some words, perhaps an attempt to scream in Internet argot or to emphasize a word?

How to Spot a Troll Bot

In Mona M’s case, it was easy. When I got the first reply I clicked over to look at Mona’s Twitter page and saw there is no description and no owner image. There is no sign of a human anywhere. If you visit Mona’s page (@Monam7M) you will see the account started in May 2016, has tweeted 1,395 times (as I write this) and has very few followers (all of the followers look like bots to me just glancing at their headings).

Also, there are a lot of tweets in one day, another indication of a bot. But not necessarily as I tweet a lot, and I am human (at least I was the last time I checked ;-)).

The biggest tip-off this is a troll bot account is all the tweets on this account are slam tweets. Most of them are directed at Canadian tweets but recently the bot has also started slamming tweets related to Clinton and sending tweets supporting Trump.

What to Do When a Troll Bot Appears

1. Do not panic (stole that from Douglas Adams, so sorry we lost him and his genius…)
2. Poking the troll the way I have can make for some fun times, but humans quickly lose interest.
3. Report the account to Twitter as a spam account.
4. You can block or mute the account if that makes you feel better.
5. Rinse and repeat as needed. There must be thousands of these things on Twitter and since all it takes to create a Twitter account is a working e-mail address, these things will be there until Twitter stops them.