In the lead-up to the election, allegations are flying that social media platforms have been “shadow banning,” or in some way filtering the political content of, celebrities. After multiple Latine stars spoke out against the racism and misogyny on display at last week’s Madison Square Garden rally, Ricky Martin posted an Instagram Story claiming that the platform was blocking a post he’d published on the subject. Users also speculated that their access to other public figures who posted about the rally, including Bad Bunny, was being somehow restricted. Add these musicians to a long list of users who say they’ve been shadow banned. These include Bella Hadid, who claimed in 2022 that the platform penalized her for posts about Palestine, and sources for a recent Washington Post investigation who claimed Instagram throttled their political content.

Claims of political post suppression have been made for years, but the idea that a platform would be censoring this content right before a deeply consequential election is alarming. Is it true?

Well, yes and no. Instagram and its text platform, Threads, have claimed in the past and continue to claim that they don’t shadow ban specific people. “Our policies are designed to give everyone a voice while at the same time keeping our platforms safe,” Meta’s Public Affairs Director Dani Lever told Vox in an email. “We’re currently enforcing these policies during a fast-moving, highly polarized, and intense election and we readily acknowledge errors can be made, but any implication that we deliberately and systemically suppress a particular voice is false.”

Last year, Instagram head Adam Mosseri detailed the many moving parts that go into determining how a post gets ranked, curated, and served to users on a platform, including, he said, “thousands” of “signals” of user preferences, including what users like, information about the post and who posted it, and how interested you are in the person who made it. This criteria varies from user to user, and it’s constantly in flux.

Instagram’s complicated system of ranking content involves another operation of filtering content as well. Previous reporting has indicated that Instagram and Threads do remove and filter out content, including covertly restricting individual accounts and their content from search results — the oldest commonly understood meaning of a shadow ban — as well as showing the posts to fewer users. That apparent contradiction may be why shadow ban discourse flared up again this past week.

In addition, the political context — mere days away from an anxiety-inducing election — adds another layer of concern. The platform’s current official policy regarding politics is not to recommend any political content to users. “Our goal is to preserve the ability for people to choose to interact with political content, while respecting each person’s appetite for it,” Mosseri said earlier this year when the policy was announced.

“People have told us they want to see less political content on Facebook,” Lever told Vox, “so we have spent the last few years refining our approach to reduce the amount of political content seen in Feed and other surfaces.”

Users can opt into being served political content if they want — which is good — but users who don’t know they need to opt in first might go looking for political content and then get the wrong idea when they don’t find it.

Not only that, but in March, Instagram rolled out a little-noticed but massive change to the way it displays hashtags across the site: It essentially erased real-time feeds altogether. Now, regardless of what tag you’re searching for, you can no longer see posts made in real time on a universal feed. While you can still see posts that are recommended for you on the mobile app, you can’t see everything. This, again, is another wrinkle that might make people think they’re being targeted or that their posts are being hidden, when they’re more likely just being lost in Meta’s ineffable algorithmic approach to sourcing content.

Given the polarized era we’re in, all of these factors can create confusion. After all, Meta has previously limited political content across all its platforms. In 2022, it issued an ad restriction in the week prior to the election to squelch all political advertising from getting through, and it’s conducting a similar restriction this week. (A recent Forbes investigation found that prior to the current week-long ban on election advertising, the company has been profiting handily from political Facebook advertising — even ads that spread rampant disinformation about the election.)

Lever pointed out to Vox that the company had actually announced its forthcoming approach to the elections a full year ago. “We very clearly stated that we would show people links to official information about how, when, and where to vote when they searched for terms related to the election on Facebook and Instagram,” Lever told Vox. Still, the average Instagram user likely won’t see missing content as part of a site-wide design; they may see it instead as an unfair ban on a specific piece of content or theme.

Users’ perceptions that their speech is being curtailed has become a running theme for Meta, as well as other platforms like Twitter, now known as X. Owner Elon Musk has frequently come under fire for allowing censorship and artificial amplification of content on that platform, often at the behest of right-wing authoritarian governments. (Vox has reached out to X for comment.) Under Musk, X has also been known to ban left-leaning users, including journalists, from the platform despite the users having violated no site policies. All of this prior activity tends to feed the rumor mill in times of high anxiety and tension — like, say, days before an election. Even if X isn’t filtering content, the perception that it must be can contribute to user backlash.

The problem of users believing they’ve been shadow banned reflects larger tensions between tech and politics

That leads us to another, perhaps more ominous, problem. Increasingly, the prominent moguls at the helms of these platforms seem willing to bend the knee to Donald Trump. Some tech entrepreneurs like Jeff Bezos and Patrick Soon-Shiong may be hedging their bets in anticipation of a Trump election win. Musk has made it clear that he is a Trump fan, but that’s not necessarily the case with Mark Zuckerberg, with whom Trump has had a famously tense relationship over the years.

Following the January 6 insurrection, Facebook and Twitter both banned Trump from their platforms, which left the former president thoroughly disgruntled, with Zuckerberg in particular. That March, on TruthSocial, he indicated his intent to protect TikTok in the face of attempts by US politicians to ban the platform — solely because TikTok is Facebook’s biggest competitor. Trump followed that up by threatening to put Zuckerberg in prison if he filtered Trump-related content from Facebook.

Trump’s statements are in line with a long litany of threats he’s made against his enemies across politics, the media, and Silicon Valley if he were to win office. Additionally, Congress has shown a willingness to investigate social media platforms when they’re displeased by alleged content restrictions.

The tensions between Trump and Meta seem to be thawing, however, as Meta has been giving progressively more face to Trump. Zuckerberg restored Trump’s access to Facebook and Instagram in 2023 after a two-year ban. He also praised Trump for his bravery following a July assassination attempt, and called him after the incident. While Trump had previously referred to Zuckerberg as an “enemy of the people,” that vibe has shifted; on a recent podcast interview, Trump claimed he likes Zuckerberg “much better now.” Facebook has denied Trump’s claim that Zuckerberg went one further and implied he would be voting for Trump in the election.

It’s not hard to see that in a scenario where tech CEOs are falling in line to pay their dues to Trump in case he wins the election, the end user is the one who loses out.