Palestinians’ digital rights ‘violated’ by censorship on Facebook, Twitter, and Instagram, new report claims

Exclusive: Users trying to share information from conflict zones saw posts taken down without explanation, through moderation policies campaigners say are ‘clearly biased’

(REUTERS)

There has been a “dramatic increase” in the censorship of Palestinian political speech on social media over the past two weeks, during the period of intense fighting between Israel and militants in Gaza.

Facebook, Twitter, and Instagram have all been used by Palestinians to share information from, among a variety of areas, the East Jerusalem neighbourhood of Sheikh Jarrah where families face eviction.

However the report from 7amleh, The Arab Center for the Advancement of Social Media, shared exclusively with The Independent, argues that social media companies’ moderation attempts and codes of conduct have resulted in numerous citizens’ accounts being taken down.

It comes in the context of huge criticism surrounding the Israeli government’s military decisions, which include displacing 52,000 Palestinians via air strikes, causing the deaths of numerous children, bombing the Associated Press and Al Jazeera building, and, on social media, the bizarrely flippant tone of its Twitter account.

Overnight, Israel and Hamas have since entered a “mutual and simultaneous” truce, after Israel’s security cabinet agreed to put an end to heavy bombardment which has killed more than 230 Palestinians.

Twelve people have been killed In Israel, including two children and a soldier. The Israeli military said 4,340 rockets were fired at Israel by militants over the course of the 11 days of fighting.

It is unlikely, however, that this will be the last time the conflict rears its ugly head, or that social media companies’ moderation decisions will not exacerbate future battles in the region – as seen by their long-ranging and concerning approaches to Palestinian content in the past.

7amleh documented 500 cases of what it calls the digital rights violations of Palestinians between 6 May and 18 May this year through a form shared via its social media channels with the support of partners including MPower Change, Adalah Justice, Jewish Voice for Peace, and Eyewitness Palestine. These violations include content being taken down and accounts being removed or their visibility restricted.

Half of the 500 instances were on Instagram, the report states, with 179 cases on its parent company’s platform Facebook; Facebook also apparently increased geo-blocking, where social media companies target the geographical location of content to help their moderation efforts, with “a number of these cases [documented] for activists from the occupied Palestinian territory”.

The organisation states that 45 per cent of all reported violations on Instagram were due to deleted Stories, with users receiving no prior warning or notice. While Instagram did not respond to 7amleh about 143 of the cases submitted, it confirmed that “only one case violated the community standards”. Instagram admitted the removal issues on 7 May, but 7amleh says the majority of reports (68 per cent) occurred after the problem was seemingly addressed.

As well as these holistic problems with content moderation, there have been specific, dramatic cases of harmful flaws in the company’s content moderation, such as Instagram removing or blocking posts with hashtags for the Al-Aqsa Mosque, the third-holiest site in the Islamic faith, as its moderation system mistakenly deemed the religious building a terrorist organisation.

“We know there have been several issues that have impacted people’s ability to share on our apps, including a technical bug that affected Stories around the world, and an error that temporarily restricted content from being viewed on the Al Aqsa hashtag page. While these have been fixed, they should never have happened in the first place”, Facebook told The Independent in a statement.

“Our policies are designed to give everyone a voice while keeping them safe, and we apply these policies equally, regardless of who is posting or their personal beliefs”, the company added. “Our dedicated team, which includes Arabic and Hebrew speakers, is closely monitoring the situation on the ground”. It added that it was continuing to review 7amleh’s reports.

There have also been instances of Facebook blocking the accounts of Palestinian journalists, a critique which has been also levied at Twitter – on which there were 55 cases of violations of Palestinian content, 91 per cent of which were suspension of accounts, according to 7amleh.

“Our automated systems took enforcement action on a number of accounts in error by an automated spam filter. We are expeditiously reversing this action to reinstate access to the affected accounts, many of which have already been reinstated”, Twitter said in a statement, adding that it had an appeals process for such accounts.

Twitter also temporarily restricted the account of Palestinian-American writer Mariam Barghouti, who was reporting on Palestinians being evicted from Sheikh Jarrah. "We took enforcement action on the account you referenced in error. That has since been reversed," Twitter said in a statement, changing Barghouti’s account to say that it was “temporarily unavailable because it violates the Twitter Media Policy.”

Twitter said that “if an account’s profile or media content is not compliant with our policies, we may make it temporarily unavailable and require that the violator edit the media or information in their profile to comply with our rules. We also explain which policy their profile or media content has violated.” Twitter did not explain to The Independent which policy was violated.

The issues around social media moderating content, specifically about Israel’s war against Palestine, are long-running. In 2016, the Israeli government announced a formal collaboration with Facebook’s Tel Aviv office, that were “meant to force social networks to rein in content that Israel says incites violence”.

Internally, Facebook listed globally protected “vulnerable” groups including homeless people, foreigners, and Zionists – a person who supports the re-establishment of and support for a Jewish state in the Holy Land, currently located in Palestine - in documents revealed by The Guardian in 2017.

In January 2021, Facebook apparently proposed a revision of the term ‘Zionist’ that would make it a proxy for ‘Jew’ or ‘Jewish’, although the company said that no decision had been made. An anonymous Facebook moderator who spoke to The Intercept said the policy, in practise, “leaves very little wiggle room for criticism of Zionism”.

That decision was criticised by Rabbi Alissa Wise, Deputy Director at Jewish Voice for Peace, who said that restricting the word “prevent[s] its users from holding the Israeli government accountable for harming Palestinian people.”

Facebook said it “understand[s] that the word ‘Zionist’ is frequently used in important political debate. ... that’s why we allow critical discussion of Zionists, but remove attacks against them in specific instances when context suggests the word is being used as a proxy for Jews or Israelis, both of which are protected characteristics under our hate speech policies.”

It added: “We always work to apply our Community Standards as accurately and consistently as possible, and don’t remove content that doesn’t break our rules. We have a clear process for handling requests from governments and regulators, which is the same around the world. We’re public about how many pieces of content we restrict locally for breaking local law, and publish these numbers in our Transparency report twice a year.”

However, much like how western news media headlines have slanted towards pro-Israeli language, reflecting the foreign policies of their national governments, many of the policies put forth by American companies are informed by US culture and norms, Jillian York, Director for International Freedom of Expression at the Electronic Frontier Foundation, told The Independent.

She added that the “US has historically been a strong supporter of Israel and has long dehumanized Palestinians, so it isn’t surprising that corporate policies would align with that worldview”.

For Facebook “it’s more important to censor terrorism than it is to ensure that Palestinians can speak freely”, York continued. “Amidst the pandemic, this has only gotten worse, as content moderators in some countries are still stuck at home. As such, we’re seeing more ‘bugs,’ more keyword filtering, more shadowbanning and other subtle enforcement tactics that aren’t simply ‘takedowns’.”

It is for this reason that social media is so important in reframing the conversation of the Israeli-Palestinian conflict, allowing messages that may not come through traditional media to reach the general public.

Snapchat’s Snap Maps feature, for example, has been used to demonstrate in real-time the difference in the effects of the conflict between Tel Aviv and Gaza, as have TikTok memes.

Both companies present a challenge to Facebook and Twitter but, York says, the network effect of these companies is still hard to disrupt and when it comes to “bold” decisions, such as banning President Trump, many follow one another.

“Facebook's rules related to Israel-Palestinian have always been opaque and one-sided. Marwa Fatafta, a policy manager at Access Now, told The Independent.

“It's no secret that Facebook often bows to government pressure and converts such demands into rules governing online speech. But that’s only half of the story [as] social media platforms rely on algorithms to moderate speech at scale and being blind to context as they are, lots of legitimate content get flagged and taken down.” Such issues stress the need for algorithmic transparency, which Fatafta says is “clearly biased”.

At some big technology companies, employees are very conscious of this power. This week both Jewish Google workers and Apple staff have called on their respective executives to recognise that “millions of Palestinian people currently suffer under an illegal occupation,”.

In affecting change, 7amleh recommends a number of practises to improve social media companies’ moderation in the end of its report.

These include hiring fact checkers specifically for Israeli and Palestinian content, allowing people access to geo-spatial information needed to respond to the humanitarian crises, providing transparency on voluntary takedown requests, and conducting human rights assessments “that includes the impact of Israel on Palestinians in Israel and the occupied Palestinian territory”.

Censorship and bias have been issues for years, however, and the escalation of violence over the past two weeks “only scaled it up and made it more pronounced”, Fatafta says.

“Social media has been a life-line for Palestinian activists deprived of access to mainstream media, and the despite of the ceasefire, the reality of occupation and oppression continues. So Palestinians will continue to use social media to organise and dissent. The main question here is: would social media companies learn their lessons this time?”