Online Hate Index is Frightening


It is reported that Twitter, Facebook and YouTube are using an online hate index created by the Anti Defamation League to moderate hate speech on social media. The online hate index is a machine based AI programme that searches out and identifies examples of hate speech so that action can be taken. This opens a massive can of worms concerning what hate speech actually is and what action should be taken against it.

In this video produced by the Anti Defamation League, examples of subjects related to hate speech include ‘the Donald’ ‘Men’s Rights’ ‘Guns are cool’ and ‘White Rights.’

While it is possible that any conversation containing these subjects could contain examples of hate speech, it is telling that these are the sort of areas where the programme is taught to look. Any programme is only as good as the people programming it and, in this case, it would appear that the programmer’s political bias is apparent in the areas it has taught the programme to search out. Assuming that anyone using the phrase ‘the Donald’ may be associated with hate speech is merely assuming that it is your political enemies who will be engaging in hate speech. It is as if the programme has been asked to target rednecks to see if they are engaging in hate speech.

There are no examples in this video of what hate speech actually is. The problem with hate speech as a concept is that it can quickly come to mean someone disagreeing with your deeply-held political views. The subject of ‘Men’s Rights’ would appear to have no reason to associate with hate speech unless hate speech is defined as disagreeing with feminism. Similarly the subject ‘Guns are cool’ should not be a repository for hate speech unless hate speech is defined as disagreeing with gun controls. The most worrying thing from the video is that it appears the only place the programme is looking for hate speech is in conversations using phrases adopted by those on the political right.

It is as if the liberal bias of the programmers and ultimately the owners of social media platforms is leading them to root out the Right through the medium of attacking hate speech rather than hate speech being the ultimate issue.  Shutting down the accounts of anyone engaged in conversations touching on the topics outlined in the video would appear to be an attack on the Right rather than an attack on hate speech, whatever that is.

This subject also raises the issue of free speech and freedom of expression. The US has some of the best protections for free speech in the world. To what extent are social media platforms undermining free speech by searching out what they determine is hate speech to shut it down? This will be a matter of degrees. There is some speech that is so offensive that social media platforms have a right to remove it, when it crosses the line into illegal speech. Who will be the judge of when that line is crossed?

Is expressing hatred now prohibited speech in and of itself? Certainly in the real world expressing hatred is still protected speech in many circumstances. Is the online world holding up a different standard? To what extent is the ability to express oneself on social media becoming a right that should be protected against the owners of social media platforms stifling free expression? These questions are far more complex than they may appear at first reading.

The impulse that drives the owners of social media platforms to turn to programmes like the online hate index are probably good ones. The desire to protect people against harassment and being exposed to speech that will cause deep offence. The problem is that in seeking to protect one group of people the owners of social media platforms run the risk of trampling on the rights of another group of people.

Free speech is protected for good reason. A free society is one in which all of its citizens are free to express their views, however offensive they may be to others. As our society’s discourse takes place more and more online the test of whether speech is free moves online also. It would appear from the Anti Defamation League’s video that there is a real risk that free speech is being interfered with due to political bias. This can never be right. It is an area to watch closely in the future.