A guy's Twitter account got suspended after he made a death threat—against a mosquito

Twitter wants you to know it's taking online abuse very seriously.

So much so that it's blocked a user who made death threats against a mosquito.

SEE ALSO: Let's all take a deep breath and stop freaking out about Facebook's bots 'inventing' a new language

A Japanese user sent out a furious tweet after he was bitten by a mosquito, saying "Bastard! Where do you get off biting me all over while I'm just trying to relax and watch TV? Die! (Actually you're already dead)."

He later received a notice from Twitter saying his account had been permanently frozen and could not be reactivated.

According to news outlet SoraNews24, this was the message the user received: 

"Thank you for using Twitter. 

Your account has been frozen because it was used to send messages containing threats. 

Tweets containing threats are not allowed under our terms of service.This account cannot be reactivated.

Thank you for your understanding."

The Japanese user, whose original account was named @nemuismywife, later created a new account @DaydreamMatcha, to rant about Twitter's decision.

24 Aug

Twitter Japan 

 ✔@TwitterJP

Replying to @TwitterJP

#夏のTwitter豆知識　表現の自由を尊重していますが、安心してさまざまな意見や信念を表現できるよう、嫌がらせや脅し、または恐怖を与えて他の方が発言できないようにするような攻撃的な行為を禁じています。
【その2】攻撃的な行為https://support.twitter.com/articles/253501 pic.twitter.com/C42W5Dbztx

 Follow

Sice @DaydreamMatcha

蚊を殺害したら前のアカウントが永久凍結されました、これは違反行為でしょうか？
Twitterは徳川綱吉ですか？ pic.twitter.com/bNh4pI8pTb

5:11 AM - Aug 26, 2017

•  117117 Replies
 
•  33,21733,217 Retweets
 
•  28,90228,902 likes

Twitter Ads info and privacy

"My account was permanently frozen after I said I killed mosquitoes...this is a violation?" he said in a tweet addressed to Twitter Japan.

Cyberbullying and online abuse is a huge problem for Twitter, with the social media platform recently announcing that it was "taking action" against abusive accounts "10 times more" than it was at the same time last year. 

Twitter also last year debuted an algorithm to detect abusive behaviour — which is probably how @DaydreamMatcha got blocked in the first place.

Artificial intelligence scripts go about looking and picking up key words, which in this case could be a combination of "Bastard", "Die" or "Dead." 

It's unlikely that a human administrator reading the whole tweet and understanding the context of it would report the same tweet for online abuse. 

Looks like AI still has some way to go.

[H/T: SoraNews24]