Facebook, Hate Speech, and the Need for Transparency

Last week, Reuters reported that Facebook and Instagram parent company Meta Platforms had quietly revised its global hate speech policy in a series of communications to its content moderators. Under longstanding policy, threats of violence and hate targeting specific nationalities have long been banned on Facebook, but under the new guidelines, the company would now permit “violent speech that would otherwise be removed under the Hate Speech policy when: (a) targeting Russian soldiers, EXCEPT prisoners of war, or (b) targeting Russians where it’s clear that the context is the Russian invasion of Ukraine (e.g., content mentions the invasion, self-defense, etc.).”

The policy also allowed for calls for the death of Russian President Vladimir Putin or Belarusian President Alexander Lukashenko, except where such calls included additional targets or information suggesting they were a credible threat, such as mentioning a specific location and murder weapon. Facebook similarly permitted calls for the death of Iran’s supreme leader last year.

While the new guidance states that the company’s “Hate Speech policy continues to prohibit attacks on Russians,” the resulting confusion forced the company to clarify that it was now “narrowing its focus to make explicitly clear in the guidance that it is never to be interpreted as condoning violence against Russians in general” and that “there is no change at all in our policies on hate speech as far as the Russian people are concerned. We will not tolerate Russophobia or any kind of discrimination, harassment or violence towards Russians on our platform.”

After formal protest by the Russian government, the company also clarified that “we also do not permit calls to assassinate a head of state . . . So, in order to remove any ambiguity about our stance, we are further narrowing our guidance to make explicit that we are not allowing calls for the death of a head of state on our platforms.”

According to Reuters, the new policy initially applied to users in Armenia, Azerbaijan, Estonia, Georgia, Hungary, Latvia, Lithuania, Poland, Romania, Russia, Slovakia, and Ukraine. Within a day of the policy becoming public, however, the company narrowed its application to users in Ukraine.

Earlier, during the invasion’s opening days, Meta had quietly relaxed its ban on praise for Ukraine’s Azov Battalion, which had been on the company’s list of prohibited organizations –  alongside the Islamic State and Ku Klux Klan – responsible for “serious offline harms” and “violence against civilians.” The battalion has neo-Nazi roots, with its founder suggesting that Ukraine’s national purpose was to “lead the white races of the world in a final crusade … against Semite-led Untermenschen [subhumans].” Yet, the battalion’s role in defending Ukraine during the invasion led Facebook to permit selected praise for its members.

Meta developed and deployed these policies entirely in secret, with the only public awareness of these shifts coming through internal leaks. The company historically has touted its reliance on outside experts in the development of its policies, raising the question of how extensively it vetted its latest policy changes – especially given its failure to name any outside organizations or experts endorsing them.

Meta’s recent moves raise questions as to whether it will apply similar exemptions to other conflicts around the world. In justifying its new policy, the company argued that “If we applied our standard content policies without any adjustments we would now be removing content from ordinary Ukrainians expressing their resistance and fury at the invading military forces, which would rightly be viewed as unacceptable.” Yet, Palestinians have long argued that their own lands have similarly been militarily invaded and occupied by force – claims that Facebook, with its AI-powered moderation systems, has systematically silenced, as the company has acknowledged.

Under its rationale for Ukraine, would Meta now permit Palestinians to call for the death of those whom they view as Israeli occupiers, or permit praise of Hamas? In a future Afghanistan or Iraq-like conflict, will Meta permit calls for violence against U.S. military personnel and allied civilians?

Domestically, a growing number of U.S. municipalities and universities are publishing statements acknowledging that the land upon which they stand is former tribal land that was forcibly taken through invasion and occupation. Would Meta permit calls for violence against such entities – or against the U.S. government itself?

Asked such questions, the company did not respond.

It appears that Facebook’s external oversight board was not involved in the development of its policy changes. The company has noted internally that it would soon be forwarding its materials to the board for post-mortem review.

That Facebook felt it had to formally change its policies to allow Ukrainians to speak out against what is happening in their country underscores the challenges of enforcing a single set of speech rules to govern the entire world. The company’s continual revisions and clarifications of the rules once they leaked publicly offer a reminder of the power of transparency in forcing companies to address the consequences of their policies.

If, in exchange for keeping their federal Section 230 protections – which immunize them from liability for hosting third-party speech – social media platforms must make their moderation rules and decisions more transparent, then Facebook won’t have to change its rules next time around just to allow users to condemn the military invasion of their own country.