The man now on trial for the 2018 Tree of Life massacre, which killed 11 worshippers and wounded six more, entered the Pittsburgh synagogue with an agenda and an ideology. Before walking in with semi-automatic weapons and handguns, he was part of an online community rife with antisemitic conspiracies, extremism and white supremacy.
What’s online and what’s offline are no longer separate.
Academics, jurists and tech experts have debated for years whether internet companies and social media platforms should be liable for harmful actions spurred by hate speech, including antisemitic messages, that they carry or even promote to their users via algorithms.
In May, the Supreme Court had two opportunities to weigh in on a crucial protection for social media companies known as Section 230 — and decided to leave it intact. The decisions gave the tech industry a reprieve but left open the possibility of congressional action.
“I think what they’re saying is: This is an issue for Congress. Congress is the one who writes these laws, and Congress should be setting the policy for internet speech,” said Aaron Mackey, director of litigation for free speech at the Electronic Frontier Foundation, a San Francisco-based nonprofit that supports Section 230.
The high court unanimously ruled last month against holding Twitter liable for permitting ISIS to upload content and then letting algorithms recommend it. The justices then declined to rule in another case involving Google and ISIS content. Both cases hinged on Section 230 of the federal code based on the 1996 Communications Decency Act, which basically states that internet companies aren’t responsible for their users’ speech.
By protecting website hosts against civil liability, Section 230 allows them to function without a crippling onslaught of civil lawsuits against speech of all kinds, Mackey said. (Section 230 does not protect them in the case of criminal acts.)
But others say Section 230 gives online platforms too much immunity for the kind of hate speech, including antisemitism and white supremacy, that can foster actual acts of violence like the Pittsburgh massacre.
Lauren Krapf is policy director and lead counsel at the Anti-Defamation League’s Center for Technology & Society, which focuses on fighting online bias. Court interpretations of Section 230 so far have given internet companies “near blanket immunity” from the consequences of hate speech on their platforms, she said.
“This immunity shield has created a lack of incentive for platforms to really put first the safety and security of users,” she said.
Section 230 was written in the mid-1990s, a vastly different internet climate from today. Internet access was slow, limited and wired. Online forums existed, but social media did not. Designed to promote innovation and encourage a free market, Section 230 identified internet platforms as distributors of speech — not as publishers in the sense of traditional media — and thus protected them from civil suits. “No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider,” it states.
Who decides what speech is out of bounds? And who’s going to enforce it?
But the landscape has radically changed. We live in a completely different digital world, with instant, constant and wireless access and where social media dominates people’s days.
Section 230 isn’t “realistic in light of the way that the internet and social media platforms specifically operate today, which is really different than the way that [they] operated in the mid-1990s,” Krapf said.
Krapf also pointed out that there’s a big difference between hosting harmful content and promoting it. Today, internet companies in general and social media networks in particular use algorithms that push “engaging” content. But it’s been repeatedly documented that “engaging” can mean “controversial” or even “extreme” content. Such content can radicalize people and take some of them down dark paths that lead to violence.
The use of algorithms makes internet companies more than passive hosts, Krapf said.
“It’s not just content that’s posted up on a platform. It’s content that’s then fed into recommendation engines, that’s pumped to the top of our news feeds,” she said.
She added that the ADL doesn’t support holding platforms responsible for all user speech. But right now, she said, the companies are protected “whether their recommendations are discriminatory in nature or whether they auto-generate content that might be related to terrorist content or might incite violence.”
ADL wants stricter rules, although Krapf said the organization’s position is that Congress, not the Supreme Court, should be responsible for toughening Section 230.
Mackey has a different view.
“I think 230, in popular consciousness, is seen as this law that it sort of enables the bad behavior of individuals,” Mackey said.
Mackey added there’s another part of Section 230 that often gets overlooked in these debates. It also provides protections that allow internet companies to remove content they find objectionable, without getting sued. The provision defines it as “material that the provider or user considers to be obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable, whether or not such material is constitutionally protected.”
“What’s important for people to understand is that [Section 230] also enables the moderation of that awful speech and gives the platforms the power and choice to decide,” he said.
He noted that this is the basis for a lot of the internet’s functionality.
“It also enables us to create sort of ‘niche’ sites that are dedicated to particular viewpoints,” he said. “So there [are] forums about veganism, forums about how to keep kosher.”
Aaron Terr, director of public advocacy at the Philadelphia-based Foundation for Individual Rights and Expression, an organization that promotes free-speech activism, supports Section 230 and is opposed to putting the power to regulate internet companies in the government’s hands.
“Whenever you’re enforcing any kind of speech restriction, the question is always: Who decides?” he said. “Who decides what speech is out of bounds? And who’s going to enforce it?”
He said that should worry both liberals and conservatives.
“Joe Biden is going to have a very different idea of what hate speech is than Donald Trump,” he said. “Nobody on either side of the political spectrum should want their ability to speak to be left at the mercy of whoever happens to be in power at a given moment.”
However, Krapf said leaving it up to internet companies and social media platforms to self-regulate isn’t good enough, given what’s at stake.
“We think that, at this point, government has to step in and implement different regulatory tools and frameworks that can really help us move forward in this era, where our online lives are just an extension of our offline lives,” she said.