Inside Meta, a debate over when the word ‘Zionist’ is hate speech

Inside Meta, a debate over when the word ‘Zionist’ is hate speech

Meta is debating whether to more aggressively remove some social media posts containing the words “Zionist” to counter a surge of antisemitism online, setting up a potential clash over censorship during the Israel-Gaza war, according to people familiar with the private deliberations and internal guidance seen by The Washington Post.

The social media giant has told some civil society groups that it’s considering expanding how it enforces its ban against hate speech to include more uses of the term, especially when it might appear as a hateful substitute for “Jews” or “Israelis,” said the people who spoke on the condition of anonymity to discuss internal matters.

“Whether the culprit is a hardened antizionist or a White nationalist, the term ‘Zionist’ often is used as an ugly synonym for ‘Jew,’” said Jonathan Greenblatt, CEO and national director of the Anti-Defamation League, who added he had not been consulted on Meta’s potential policy changes.

Antisemitism has soared on social media platforms in the wake of the Oct. 7 Hamas attack on Israel and Israel’s subsequent campaign in the Gaza Strip, a bloody conflict that has now claimed tens of thousand of lives and displaced the vast majority of the population. On platforms including Meta’s Facebook, Greenblatt said, “antisemitic posts … directly invoke the words ‘Zionist’ or ‘Zionism’ in a manner that is not just pejorative but antisemitic, threatening and shameful.”

Antisemitism was rising online. Then Elon Musk’s X supercharged it.

But while Meta’s move may appeal to Jewish groups that have long accused the company of being slow to address antisemitism, it has triggered alarm among digital rights activists and pro-Palestinian groups, who say the approach would stifle legitimate political critiques of the Israeli government, its armed forces and Zionism during a catastrophic war.

“Zionism is an ideology. It’s not a race,” said Nadim Nashif, co-founder of the pro-Palestinian digital rights group 7amleh, who was briefed by Meta on the policy review. “As I told them, in my opinion, this is a slippery slope. From there, you can remove a lot of content that is criticizing Israel and Zionism that is part of legitimate political discussion.”

Meta spokeswoman Erin McPike said in a statement that the company doesn’t allow users to attack people on the basis of religion or nationality but that the company needs to understand how people “use language to reference those characteristics.”

“While the term ‘Zionist’ often refers to a person’s ideology, which is not a protected characteristic, it can also be used to refer to Jewish or Israeli people” themselves, McPike said. “Given the increase in polarized public discourse due to events in the Middle East, we believe it’s important to assess our guidance for reviewing posts that use the term ‘Zionist.’”

Under current rules, Meta bans attacks people based on race, religion, nationality or sexual orientation. The company also may take down posts that spread “harmful stereotypes,” curse or generally dehumanize groups of people.

Growing Oct. 7 ‘truther’ groups say Hamas massacre was a false flag

Meta has long considered most content discussing political ideologies, governments or institutions fair game. However, it will remove some posts criticizing Zionism when it finds the term being used as a hateful synonym for Jewish or Israeli people. For instance, according to internal guidance obtained by The Post, Meta currently removes such statements as: “This city is full of Jews. I hate those Zionists.” “Zionists are a bunch of rats.” And “Kill the Zionists.”

Now, Meta is evaluating whether to expand enforcement to include posts where the word is less obviously used as a slur, the people said. In one hypothetical exchange under review by Meta, a user says, “If the media are attacking you, you’re doing something right” and a commenter responds, “Just say it, THE ZIONISTS are manipulating you.”

In that example, debate is focused on whether the commenter’s intention might be to spread a harmful stereotype about Jewish people controlling powerful institutions, the people said.

Other hypothetical phrases under evaluation for more aggressive enforcement, according to the guidance: “Zionists are war criminals, just look at what’s happening in Gaza.” “I don’t like Zionists.” And “No Zionists allowed at tonight’s meeting of the Progressive Student Association.” Under current policy, those posts might be removed if they referred to “Jews” or “Israelis,” instead of “Zionists.”

Historically, Meta has often preferred a hands-off approach to moderating content despite pleas from civil advocacy groups of all types to take down more offensive content. Meta CEO Mark Zuckerberg defended users’ right to post Holocaust denialism online in 2018 — a year after a deadly neo-Nazi rally in Charlottesville. The company changed its policy in 2020 after years of criticism from Jewish groups.

Pro-Palestinian creators use secret spellings, code words to evade social media algorithms

The Israel-Gaza war has sparked a fresh round of condemnation against tech companies for failing to catch hateful content. ADL said the group has received more complaints about antisemitic posts on Facebook and Instagram than any other social platforms, but that Meta has taken action on only 23 percent of flagged posts.

“This is an embarrassing level of responsiveness,” Greenblatt said.

Meanwhile, pro-Palestinian groups have long argued that Meta and other tech companies have failed to protect them from hate speech while suppressing legitimate critiques of Israeli policy. Last fall, throngs of Palestinian supporters complained that Meta was suppressing their view counts and video likes on Facebook and Instagram as they commented on violence in the region. At the time, Meta blamed a bug for preventing some posts, ephemeral videos known as Stories and short-form videos known as Reels from showing up properly, but said the bug had affected accounts equally around the globe, regardless of content.

It wasn’t the first time a glitch affected content in the region. During a two-week war between Israel and Hamas in 2021, Israeli police stormed the al-Aqsa Mosque, a sacred Muslim site in Jerusalem, prompting Hamas to fire rockets into Israel. Israel retaliated with a bombing campaign that left more than 200 Palestinians dead. Instagram restricted content containing the hashtag #AlAqsa — a glitch Meta initially blamed on an automated software deployment.

Israel-Gaza war sparks debate over TikTok’s role in setting public opinion

Later, however, an outside audit commissioned by Meta found that the #AlAqsa hashtag had been mistakenly added to a list of terms associated with terrorism by a third-party contractor. The report noted that Meta’s systems, reliant on artificial intelligence, were more likely to flag Arabic content as being associated with terrorist groups.

Earlier this month, Palestinian digital rights activists with 7amleh pressed Meta and other companies to try to stop users on Facebook and Instagram from sharing “genocidal” statements and dehumanizing content about Palestinian people. After the International Court of Justice ordered Israel to do more to prevent the killing of civilians in Gaza, “We expected that Meta would come to us … and they would say, ‘Yes, we have to make corrections on our website to protect Palestinians,’” Nashif said.

That didn’t happen, he said, and now Meta’s policy review threatens to further silence Palestinian voices.

“Offering a political ideology protection … sets a dangerous precedent for freedom of expression online,” said Marwa Fatafta, a policy and advocacy director for the digital rights group Access Now. “To distinguish between legitimate criticisms of Zionism and disguised antisemitic attacks requires nuance that neither Meta’s algorithms nor their overworked content reviewers can get right.”

“It will result in Meta policing more speech,” Fatafta said, “and marginalized voices will be the first to suffer.”