On one side of this fraught moment: the president of the United States, facing multiple crises less than six months before the election. On the other: Twitter, the social media giant, which has grappled for years with how to handle its most prominent — and divisive — user. Caught in the middle: reality itself, and whose version gets heard over all the noise.

Twitter’s decision this week to stand up to President Donald Trump by attaching warnings to some of his many tweets has been years in the making, a culmination of American divisions playing out and being amplified across social media. It is fueled by some of the very elements that make modern American discourse so polarized, so fast-moving and — at the oddest of historical moments — so fragmented.

Twitter’s assumption of a stronger referee role in its approach to Trump’s tweets reflects a “pretty radical change,” said Josh Pasek, an associate professor at the University of Michigan’s Institute for Social Research. “We really haven’t been at a place where social media companies were willing to take on this role.”

But it also heightens the dangers of polarization. “When you can’t agree on the state of the world, you open up opportunities for people to question the motives of others,” Pasek said. He says that makes existing conflicts worse and de-legitimizes people with diverging views. “You make it easier to see those people who differ from you as less American.”

For years, since long before he was president, Trump has used Twitter as a personal megaphone to build his personal brand, appeal to his supporters and attack his rivals of the moment. In the process, regardless of the facts at hand, he often creates his own version of reality — from birtherism to climate-change denial to exaggerations about voter fraud.

On Wednesday, Twitter pushed back in the mildest possible way. For the first time, it added fact-check links to two Trump tweets about the supposed risks of voting by mail.

Almost immediately, Trump turned the might of the U.S. government’s executive branch on the company that feeds and amplifies him.

On Thursday, he issued an executive order intended to chip away at the legal protections companies like Twitter enjoy. Undeterred, Twitter responded early today by flagging another Trump tweet — one that suggested he would have the National Guard fire on protesters in Minneapolis — with a warning for “glorification of violence,” which is against its rules. Users can still see the tweet by clicking through the warning.

Trump responded by having the White House Twitter account — a public account that represents the executive branch of the government — post the same tweet. Twitter soon added a warning to that as well. Hours later, Trump sought to walk back the meaning of the original tweet, writing that he had merely “spoken as a fact” that looting can be followed by civilian shootings.

The feud with Twitter serves as a convenient distraction for Trump from major challenges he currently faces, such as controlling the coronavirus pandemic and managing an economy hard-hit by COVID-19 restrictions. Unlike in those arenas, where researched hard numbers can mute the president’s points, in the me-vs.-them battle of Trump and Twitter, his ability to make his point can be based largely on volume and bluster.

“He can make hay about it,” said Melissa Ryan, CEO of the consultancy group Card Strategies, which researches online disinformation and right-wing extremism. “The administration is desperate to get the focus on anything that’s not 100,000 people dead.”

But it also places social platforms in a tough spot: to police or not to police content. Either choice is extremely risky.

The social media companies and their critics are arguing over when and how they should regulate the content on their platforms as coronavirus misinformation swirls and the 2020 U.S. presidential election looms. It has reached the point where it’s virtually impossible for social platforms to remain neutral — and where even fact-checking can mean taking a political stance.

Facebook did not touch the same posts that Twitter labeled, a position backed directly by its CEO.

“I just believe strongly that Facebook shouldn’t be the arbiter of truth of everything that people say online,” Mark Zuckerberg told Fox News on Thursday, a statement he frequently repeats at moments like these. Facebook has long used fact checks on its site, done by third-party news organizations such as the Associated Press.

Not long afterward, Twitter CEO Jack Dorsey tweeted that Twitter will “continue to point out incorrect or disputed information about elections globally.” But he added: “This does not make us an ‘arbiter of truth.’”

“Facebook’s approach is a strategic one from a business perspective,” said Dipayan Ghosh, co-director of the digital platforms and democracy project at Harvard’s Kennedy School. “Twitter is thinking about democratic interests and its impact on the world.”

The tensions will likely only escalate as November nears and increasing numbers of potential voters argue it out — and choose their facts — on the major social platforms.

Twitter’s actions add to ongoing outrage about the alleged suppression of conservative voices on social media, said Cayce Myers, an associate professor of communications at Virginia Tech. He said that risks further dividing social media platforms into echo chambers as the companies, not just the content they host, become politicized in the public mind.

“The problem in it for Facebook is (that) staying out of it is also a political position,” Myers said. “You’re on a side. There’s no way to not be on a side of this. Regardless of what they’re going to do, they’re going to be placed on a side.”