The image-sharing app Instagram is increasingly becoming a home for hate speech and hoaxes, even as its parent company Facebook works to stamp out troublesome content ahead of the midterms.
Instagram has long been viewed as free from the toxic atmosphere seen on other social media platforms, with users regularly posting photos of family, pets and travel to their personal followers.
But as the midterms near experts say a number of accounts are proliferating conspiracy theories, including about billionaire Democratic donor George Soros funding a migrant caravan headed to the U.S. and attempted bombings of prominent Democrats being a false flag.
{mosads}Facebook says it is making strides to crack down on misinformation on its own platform, but critics say the company’s efforts have forgotten about Instagram.
“Instagram right now is completely overrun with hate speech and propaganda memes. It’s possibly the worst I’ve ever seen, and only a few days before the election,” social media researcher Jonathan Albright told The Hill.
In recent weeks as the migrant caravan and pipe bombs dominated the news, Albright’s research found misinformation spreading across the platform. He tracked a growing number of posts spreading conspiracy theories that Soros was behind the caravan and controlling other movements such as NFL players kneeling during the anthem to protest racial injustice.
The conspiracy theories about Soros are rooted in far-right, anti-Semitic groups on sites such as 4Chan and Reddit, but are now appearing on an app owned by tech giant Facebook that boasts over 1 billion registered accounts.
Albright said Instagram’s search function seems to have few safeguards against sharing such content.
Users that typed in “George Soros” see suggested hashtags including “GeorgeSorosIsEvil” and “GeorgeSorosRiots” in the search tabs. Other hashtags such as “Soros” or “J3ws” – a spelling hate groups use to avoid filters – also directed users to scores of anti-Semitic content.
“Getting into people’s queries is a hugely impactful process,” Albright warned.
Other search terms led Albright to anti-Semitic content that included swastikas, praise for Adolf Hitler and references to gas chambers. That content violates not only Instagram’s community guidelines banning “hate speech” but likely German laws barring Nazi symbols, Albright noted. He said he was able to access almost all of the pro-Nazi content even after changing his IP to an address located in Germany.
“An abundance of this type of extreme content – often with ten to twenty related hashtags – has other effects,” Albright said, noting that it would affect searches and content even for “regular users.”
He said the ease in finding hate speech on Instagram was unprecedented compared to how tightly Twitter and Facebook police their content.
“At the least, extremist memes and hashtags should be put behind a mature content warning page,” he added.
More so than other platforms, the content users see on Instagram is driven by algorithms, in particular its unique “Explore” function. That feature aggregates videos and images the algorithm believes a user is likely to enjoy.
Jonathon Morgan, chief executive of New Knowledge, a security company that tracks online misinformation, said that in some cases it can mean access to increasingly divisive content.
“The Explore algorithm sees that you’re interested in anti-vaccination theories,” Morgan said. “Then it thinks, ‘you might be interested in how the earth is flat.’ Then it goes ‘oh you’re interested in the earth is flat, then you might be interested in QAnon,’” he said, referring to a hoax about special counsel Robert Mueller and President Trump.
“It’s just pulling people down rabbit holes and showing them content further and further from the mainstream.”
Experts say that misinformation and hateful content on Instagram can be particularly harmful because unlike Facebook and Twitter, users don’t expect to be confronted with partisan content on the site.
“Being exposed to hate speech, profanity, unreliable disinformation on a place that you don’t really expect to be targeted shows the vulnerability of Instagram in particular,” Albright said. “Your guard isn’t up as much.”
Instagram, despite being a massive platform, is often an after-thought in debates about internet extremism.
In recent months, major platforms have faced pressure to boot far-right groups and have responded by banning white nationalists like Richard Spencer and conspiracy theorists like Alex Jones. But Instagram has largely avoided that public pressure.
Many actors banned on other platforms are still present on Instagram. Jones, for instance, has been banned by Facebook, but still has an Instagram account with 230,000 followers to promote his content.
“I think overall Instagram has been the neglected platform both in the public conversation about misinformation, hate speech and abuse and in how such content on its platform is policed,” Morgan said.
“The conversation has focused on Twitter and Facebook,” he continued. “Both of those platforms have taken their responsibility to improve seriously. They’re not perfect, but they’ve tried.”
On every level though, he said, Instagram lags in cracking down on hate speech and hoaxes. It’s also easier to buy followers and fake engagement than on other platforms.
“It’s across the board. It’s not just on content moderation. It’s the underlying platform integrity,” Morgan said. “It feels like Twitter three years ago.”
Instagram insists that it is working to remove hate and misinformation on its platform as soon as it is detected. It also claims it has not seen a spike in hate speech.
“As with other major news stories, we’re seeing content on Instagram related to this weekend’s events. Our teams across the company have been monitoring developments in real time and how they relate to content on our sites,” an Instagram spokesperson said in a statement to The Hill.
The spokesperson noted that Instagram bars discriminatory content.
“We are actively reviewing hashtags and content related to these events and removing content that violates our policies, including hate speech,” the spokesperson added.
The company has removed some major hashtags spreading hate, including “jewsdid911,” following a New York Times story and “14888” a neo-Nazi hashtag.
In one high-profile incident, Instagram at first declined to delete a post from right-wing commentator Milo Yiannopoulos cheering the pipe bombs sent to Democrats. But after a public outcry, it reversed its decision.
On Tuesday, both Facebook and Instagram also banned the far-right group, the Proud Boys, after members of the group were arrested for attacking protesters in New York City.
Instagram’s features may also make it easier for problematic accounts to hide.
Instagram has no retweet or share feature meaning most posts pass through private direct messages or are shared within a circle of followers. That means that many accounts can sometimes lurk undetected.
Instagram’s user base is also younger on average than other platforms, which worries experts.
“The younger generations are most likely to be radicalized by content are on YouTube and Instagram,” Morgan said. “They’re more likely to get their news online and establish their views on the norms of what they see on the internet.”
Social media watchers note that Instagram was also targeted by Russian misinformation efforts in 2016. The company has disclosed accounts linked to the Internet Research Agency, a Kremlin-backed troll group. Albright’s research has also explored the extent of the influence of those operations.
Ahead of another election, Albright said it was troubling Instagram is not taking steps to clean up its platform.
“What’s happening right now on Instagram days before an election. It’s too much,” he said.