A story last week in The New York Post that appeared to show emails linked to US presidential candidate Joe Biden’s son became the center of a struggle by social media giants to limit the public’s exposure to “unsafe” news that has not been “fact-checked.”The unprecedented attempt by both Facebook and Twitter – two ostensibly rival companies that have built a collective base of more than three billion users – to stop people posting news stories has major ramifications for global media, democracy and regulators. The social media giants today influence the way most of the world receives its information and news. Most governments do not regulate these giants, so the public has little way of knowing what rights they have. Decisions appear arbitrary. For instance, this month Facebook banned Holocaust denial Why did it take so long, when other forms of hate and racism have been banned? It’s not clear. In Israel, Knesset voices such as that of Michal Colter-Wunsh have been outspoken, demanding explanations about why social media tolerate antisemitism.There are concerns about why some governments, such as Turkey, are apparently able to get social media giants to remove content, and whether other authoritarian regimes might soon be able to increase their ability to suppress critical news stories. There are also questions about why authoritarian regimes, such as Iran, can post whatever conspiracy theories it wants on social media while media in democracies, such as the New York Post, had their stories taken down or restricted.Let’s start with what we know about the story at the New York Post about Hunter Biden. Almost a week after the story broke, there is still no transparency about the decision-making process that went into suppressing the story or why this particular story violated the company guidelines. The links to several articles were targeted by social media companies.The articles claimed to expose emails linked to Biden’s son and various foreign interests. The story was scandalous and salacious. Initially, users of Twitter reported that attempts to post a link to the story resulted in no tweets being sent. In addition, attempts to click on a link to the story by those who had evidently tweeted it prior to the ban received a message that the “link may be unsafe.” The Orwellian term “unsafe” lacks clarity. Twitter said the link had been identified by its “partners” as “potentially spammy.” It might contain “malicious links that could steal personal information” or might contain “certain categories of content that, if posted directly on Twitter, are a violation of the Twitter Rules.”Twitter chief executive Jack Dorsey would later admit that efforts by the company to communicate what it was doing were not great. Direct messages of the article and blocking of the URL were both done with “zero context.” Twitter claimed the article was blocked because it “violated its private information policy and its hacked material policy,” the BBC reported. It was not clear if these policies had been previously enforced against other journalists and newspapers. Historically, media have relied on whistle blowers and others who might pass on hacked or private information.Facebook took a slightly different tack with the same story. It slowed down how the story was being shared and who was seeing it. Large social media companies rely on algorithms to decide who sees the flood of content being posted. Over the years the companies took action to limit the spread of “fake news,” and to keep users on their sites by not rewarding viral videos of things like cats that people use for click bait, and generally steal from other websites.Most of these decisions were guided by desires for profit. Politics also played a role after social media giants were accused of enabling the spread of misinformation before the 2016 US elections. The rise of ISIS forced the companies to reduce accounts linked to extremists. Now, the COVID-19 pandemic has led to unprecedented intervention, ostensibly to stop people from spreading misinformation about masks or various drugs.WHEN IT came to the story about Joe Biden’s son Hunter, Facebook said that it was reducing its “distribution” on the platform. Online publication The Intercept pointed out in an article that the company had “dispatched a life-long Democratic Party operative who now works for Facebook” to make the announcement. Facebook reportedly slows down the spread of other stories “when there are questions about their veracity, so that third-party fact checkers can have a look,” according to the BBC. There is almost no transparency about who these fact-checkers are or how they are assigned to a story, and whether a news organization can contest the “fact” checking.When it comes to politics, the suggestion that there is only one list of “facts” that can be verified is troubling because many issues are up for debate, such as whether someone accused of being “corrupt” truly is corrupt. Is it only a “fact” if the person has been convicted of some form of corruption, or is the mere appearance of corruption enough to make the accusation a “fact?” Many major historical incidents still lack clarity about how and why they transpired.The Turkish government, for example, denies the Armenian Genocide. Would Turkish government “fact checkers” be allowed to stop US readers from reading the word “genocide” in articles about the fate of Armenians during the Second World War?Recent news from France told of a museum that had to stop an exhibition on Mongolian history due to objections from China. What are the “facts” of that case? Will Beijing influence social media giants to make sure only its official “facts” are used?We don’t know how the third-party fact checkers will begin to influence what information we are allowed to see, but giving over news articles to random arbiters of truth is not likely to produce a good outcome. It is generally better for controversies to be debated openly, in the public space, rather than behind closed doors. We know a lot more about the Vietnam War, and sexual harassment in Hollywood, because people were able to come forward with information and share controversial details publicly, than had we relied on corporate-appointed fact checkers.When reports about Harvey Weinstein began to circulate years ago, journalists and news organization were put under tremendous pressure not to run stories about the developing accusations. Had fact checkers been allowed to stop distribution of the initial articles, they might have killed the story because there were no “facts” in them when the allegations first surfaced. Only later, when women were able to come forward and court cases began, were the details determined. The recent incidents at social media giants look as if they will have a chilling effect – not only on freedom of expression, but on controversial reports in general.Social media giants intervened arbitrarily in the New York Post story, and they appeared to do it in coordination. It wasn’t clear how this communication happened, but it was highly unusual that both giants would act together. However, it appears that social media giants are increasingly acting in unity to scrub their sites of certain information. For instance, after incidents of violence, such as terrorist attacks, the accounts of users linked to the incidents are often quickly deleted. The evidence of their radicalization is thus lost to the public and researchers. A woman who ran a YouTube channel and opened fire at the company headquarters in 2018 had her social media history rapidly scrubbed.YouTube and Facebook have both been deleting videos from many years ago that show atrocities committed by the Syrian regime. These are ostensibly removed because they violated some violent content guidelines. However, researchers point out that now the videos will be lost to history, and potentially lost to courts that want to investigate Damascus’s war crimes. This would be a bit like removing all the images of the Nazi war crimes and death camps simply because they show violence or nudity. That would have been convenient for the Nazi criminals, and would have left the victims in a tougher place to show what was done to them.BECAUSE GOVERNMENT regulators have not mandated that social media giants archive all their accounts before arbitrary deletions, massive amounts of evidence of crimes and the history of humanity in the last decades is vanishing at an unprecedented rate. It would be similar to burning down archives all over the world arbitrarily just because some contain offensive material. It is unclear if images from the Mai Lai Massacre committed by US troops during the Vietnam War would be permitted on social media these days. This means that jarring images from the past, whether the Holocaust or crimes in Southeast Asia or crimes of the Iranian regime, are all slowly being taken away from a generation, so they will be blind to past war crimes and human rights abuses, under the guise that the acts depicted are violent or offensive.Alongside the destruction of history that takes place during deletion of accounts, governments have not demanded that accounts be archived. There are also more frequent culling of accounts by social media giants that allege mass networks of bots. One story noted that as of 2017, there might have been 48 million “bots” on Twitter. Social media giants have often identified government-backed attempts to influence social media, including armies of fake accounts pushed by Turkey’s ruling party, and connected to other political campaigns from the US to Israel. But just deleting hundreds of thousands of accounts on a whim doesn’t really help if the evidence of how they were managed isn’t turned over to authorities and news organizations for research.The same is true regarding research into the rise of ISIS and other terrorist networks. In 2018, reports said Twitter had removed 1.2 million terrorist-linked or supporting accounts. Facebook had reportedly taken down some 14 million pages of terrorist content. This is good, but without archiving, there is no way to go back and research how social media helped ISIS grow and radicalize many its supporters. In some ways the decision of these giants to remove content also helps to hide their role in providing a platform for it in the first place.Most radicalization of terrorists in the last few decades has taken place online. This is in contrast to the 1990s, when tapes of Osama Bin Laden were passed around by extremists. Many people who have committed terrorist attacks, such as the recent beheading of a teacher in France, likely were triggered by someone seeing information online. Understanding how better to tamp down on radicalization online – perhaps by intervening with positive messages of coexistence or helping security services identify potential radicalized networks – is made less likely by deleting information that might point to a terrorist network or cell. On the other side of the coin, countries could exploit social media companies to go after dissidents by demanding information from the companies.The arbitrary actions of the social media giants have undergone little scrutiny or checks and balances from governments. One account that only provided Kurdish language lessons, for example, was arbitrarily suspended; accounts that document weapons in Iraq and Libya have been suspended; Iranian dissidents have been targeted for suspension; as have those who track planes using open-source intelligence gathering. Each time the online community has had to beg the social media giants, usually through tweets to “Jack,” to try to get the accounts reinstated. There is very little transparency as to why an account devoted to Kurdish language was removed one day and reinstated the next. It usually appears to be that the accounts were targeted, perhaps by government-backed armies of fake account-holders in places like Ankara that report the language lessons as “terrorism.”The public is often in the dark about these arbitrary decisions. Zoom, for instance, acknowledged in June that it had suspended activists’ accounts because of requests from China. Given the massive amount of influence these companies hold, the allegations that powerful governments can silence dissidents globally through a simple phone call or request, is chilling. Yet democratic governments have not bothered to demand even basic transparency regarding why some accounts are suspended and their holders silenced.THE CLAIMS that social media companies are above regulation is not logical historically. All industries that exert massive influence are regulated, from hotels to gas stations to radio networks. Trillion-dollar corporations with billions of users should never be completely unregulated. Social media giants been allowed to grow so large that they now control much of the world’s access to information. Small democracies today may only function at the whim of a corporation, and the accounts whose distribution its CEO wants to limit or enhance.We don’t know how the company algorithms work, so we don’t know the degree to which they may play a outsized role in elections. It would be a bit like handing over control of highway traffic to a company that can route the people wherever it wants, one day to IKEA, the next to the beach, moving them past advertising billboards along the way. In no other walk of life are the daily activities of so many controlled by so few.The world generally accepts the concept of social media-giant benevolence. The idea is that they only intervene when absolutely necessary to stop “unsafe” content. However, the intervention against the New York Post appears to be a turning point that went mostly unnoticed because it appeals to the left-leaning media in the US that generally feels the “Biden’s emails” story is a non-scandal, or one designed to aid US President Donald Trump.That was a good test run for the social media giants to pick an easy target so that next time a news story goes against what some people feel should be read by the public, it can be removed. Was the story stopped because of hacked emails or because it was “unsafe” or not “factual?” Was this just benevolent intervention to stop a story spreading that contained problematic information? We don’t know.Social media giants have a good reason to fear being manipulated on the eve of elections. The concept of an “October surprise” is linked to use of media to spotlight a story damaging to one candidate. In an era of globalized Internet, information spreads faster than ever before. It is this logic that guides the armies of content monitors at social media giants.But there have been many questions over the years about those content-checkers making mistakes, such as when content is flagged as unsafe or pornographic. Nudity in artworks may easily be censored under nudity policies, and anti-racism protesters may be flagged as “racist” if they post an image of a swastika drawn by vandals to raise awareness. Artificial intelligence and algorithms now used to scan information that is posted, flagging it falsely. Of equal concern, small media and non-profits have found that their attempts to pay to boost posts on Facebook were often flagged as “political” without any logic as to why. Twitter has labeled Russian government-linked accounts as state-affiliated but have not done the same with Qatari or Turkish or Iranian regime accounts. The arbitrary labeling leaves users confused.The sudden decision to limit viewership of the story at the New York Post may be seen historically as a watershed moment in social media history. News organizations and journalists covering events may now find their content, whether live streams or scandalous articles, blocked due to reasons that are unclear or arbitrary. A journalist who now receives leaks might be concerned about publishing them, lest “fact checkers” determine it is not a “fact” simply because it is new information.Those journalists who cover war crimes may find they cannot post photos of the crimes or evidence because they contain “violent content” and are “unsafe.” It is unclear when the next journalist and major media organization will be told their breaking news has been stopped from being posted, perhaps because a foreign government didn’t want it publicized. The lack of transparency in how these decisions are made has already had a chilling effect.