Home In the News Government Regulation and Stronger Algorithms Will Not Save Social Media From False...

Government Regulation and Stronger Algorithms Will Not Save Social Media From False and Hateful Speech

Sacha Baron Cohen is an artistic genius.  To many, he is also vile, duplicative, and, at times, revolting.  But he is still a genius.

As an actor, he has portrayed some dreadfully iconoclastic characters:  Borat, Bruno, and Ali G, to name three. And he has reviled many of the rather ordinary folks he has duped into appearing in his “mocumentaries.” With satiric winks to the audience, Cohen lets his participants naively cast themselves as racist, anti-Semitic, and Islamophobic fools.

His current mission, though, is to stem the tide of hate speech and false propaganda that has proliferated on the Internet.  Perhaps the antithesis of what has made him an extraordinarily successful and wealthy entertainer.

In a speech last week to the Anti-Defamation League – which honored him with an International Leadership Award – Cohen launched a scathing broadside against Facebook’s Mark Zuckerberg and other social media executives for allowing – and even encouraging – hateful speech and misleading political and social narratives. He adapted his speech into an op-ed for The Washington Post.

“Hate crimes are surging, as are murderous attacks on religious and ethnic minorities,” Cohen said.  “All this hate and violence has something in common: It’s being facilitated by a handful of Internet companies that amount to the greatest propaganda machine in history.”

His goals are well-intentioned. He rightly pointed out that fake news outperforms real news on social media.  Lies spread faster than truth.

When Winston Churchill said “a lie gets halfway around the world before the truth has a chance to get its pants on,” it was before the advent of social media.  With the speed of Twitter, a lie and circumnavigate the world several times before the truth even gets out of bed.

As proposed solutions, Cohen wants politicians and government bureaucrats to rein in social media, and he wants the social media sites themselves to bolster their algorithms to block more hurtful speech.

Neither will work.  And both proposals likely would do more harm than good.

When members of Congress cannot even agree whether it was Russia or Ukraine who hacked the Democratic National Committee for dirt on Hillary Clinton, does Cohen really think it can solve the spread of false or hateful speech through legislation and regulation?  That is folly.

There actually is legislation in the U.S. Senate to empower the Federal Communications Commission and Federal Trade Commission to create and enforce regulations to ensure that social media sites operate without “political bias.”  The proposal has gotten a frosty reception.

It’s hard to imagine that well-meaning and hard-working federal government officials would want to referee the subjective content of what is posted by average citizens on the websites of private businesses.  Or that they could be properly trained to do so. Numerous college campuses are trying very hard to regulate hate speech and finding how difficult it is.

Even more difficult is creating algorithms that would capture unacceptably vituperative speech.  There is no definitive definition of “hate speech.” And there are easy ways to trick an algorithm into believing your speech is innocent.  For instance, instead of writing “Martians are disgusting and should be killed love,” simply type “MartiansAreDisgustingAndShouldBeKilled love.”  The former almost certainly would be flagged as hate speech. The latter, with only the word “love” standing alone, would likely be deemed acceptable.

Moreover, what’s offensive to some is commonplace to others.  A scientific study demonstrated the enormous challenge – or even the impracticality – of trying to define and flag “hate speech.”

Indeed, the social media algorithms that already exist to curtail hate speech have been shown to be racially biased, chiefly against African-Americans and other minorities.

According to one study, Twitter postings by African-Americans were much more likely to be flagged as “toxic” than those by others.  Tweets written in African-American Vernacular English (AAVE) were more prone to be categorized as hate speech, in part because the use of the “n-word” is culturally more acceptable and often used in AAVE as a non-hate speech term.

Ironically, Cohen’s script for the movie Borat probably wouldn’t survive Facebook’s anti-hate speech algorithm.  It is replete with anti-Semitic and anti-Muslim language.  But, as satire and comedy, of course, it is immune from prohibition, and rightly so.  We, the audience, get the joke. But would a computer? Many think not.

Cohen is in on the joke.  He, and we, need a better solution.  Education would be a good start. Not censorship.


Robert D. Lystad has been a practicing First Amendment attorney for more than 25 years, and is the executive director of the Campaign for Free Speech.

Must Read

Are Trump’s Rally Remarks Protected by the First Amendment?

Amid the outrage and sorrow stemming from the destructive and lethal storming of the U.S. Capitol on January 6 are renewed calls from across...

When It Comes to the Coronavirus, What – and Who – Should We Believe?

If not for a free and inquisitive press, it is unclear what America would know about the health of the leader of the free...

Automation Drives Censorship on Youtube

In response to videos relating to COVID-19, Youtube, Twitter and Facebook have removed videos containing information they deemed “false” or “misleading.” For example, a...

The Many Lawsuits of Devin Nunes

Perhaps the headline writer for a Washington Post opinion column framed it best:  “Raise your hand if you have not been sued by Devin...

Don’t Cancel the First Amendment

Originally published by National Review, written by Charles C. W. Cooke. Of all the strange and fanciful notions that have bubbled up to the surface...