Description du cas
The Board will address a case regarding a post commenting on migration into and out of Ukraine including African immigrants potentially filling Ukraine’s labor shortage, by a Facebook account from Ukraine. In doing so, the Board will assess how Meta deals with contentious political topics, including in conflict settings, non-English languages and from journalists’ and political commentators’ accounts.
In April 2026, a user posted text and an image on Facebook discussing migration into and out of Ukraine and the social and economic policies that contributed to labor shortage in the country. The text, in Russian, responds to a statement made by Kyrylo Budanov, the Head of the Office of the President of Ukraine, that the country may ease its immigration restrictions to attract foreign workers, including those from African countries, to address its shrinking workforce. The user’s post ridicules public outrage in Ukraine against African migrants, noting that the war with Russia had intensified Ukraine's demographic crisis and the country may eventually need millions of migrant workers to fill the gap.
The post states that Ukraine would be lucky to have migrants “from the most backward countries,” and underscores existing racist and xenophobic attitudes toward them. It adds that the labor force problem is a consequence of the country’s poor economic policies and the disregard for Ukrainian people for decades who are “scattering all over the world like cockroaches.” The accompanying image shows a map of war-damaged Ukraine, Ukrainian soldiers, what appear to be displaced Ukrainian people, and African migrants carrying luggage. The content received about 76,000 views, 1,000 reactions and under 500 shares and comments.
Three users reported the content for Terrorism, Hateful Conduct and Fraud or Scams. Meta’s automated systems enqueued the content for review and determined that it violated the Hateful Conduct policy prohibiting dehumanizing speech “in the form of comparisons to or generalizations about animals, pathogens or other sub-human life forms,” including “cockroaches.” Following the removal, Meta applied a standard strike and time-based feature limit to the user’s account. The user appealed the content removal decision to Meta and following a review by its automated systems, the company upheld its original decision to take it down. No human reviewer assessed the content throughout the review process.
The user appealed Meta’s decision to the Board, stating that the post was intended to support migrants and draw attention to the disproportionate hostility on social media toward them. The user also noted that the post was not hate speech and he did not attack migrants based on their protected characteristics but called for more tolerance. He further argued that over the recent years he received multiple sanctions on his Facebook account and every restriction follows the same pattern of erroneous enforcement as in this case.
When the Board brought the case to Meta's attention, the company’s subject matter experts determined that the original decision to remove the post was an error. The statement that Ukrainians are “scattering all over the world like cockroaches” did not violate its Hateful Conduct policy because it was an assertion about the behavior (i.e., migration) of some Ukrainians rather than a generalization about their inherent qualities. Also, Meta determined that the statement about immigrants from “the most backward countries” is not an attack based on national origin, but a characterization of the countries’ level of development and therefore not hate speech. Meta restored the content and removed the strike. It also conducted a review of multiple strikes that the user’s account accumulated and found that it had erroneously applied approximately half of them. Meta removed these strikes from the account and restored related content.
The Board would appreciate public comments that address:
- How should social media companies design their automated detection and enforcement to distinguish political criticism and discussion about migration from attacks against migrants based on protected characteristics, particularly when enforcement does not involve human review.
- Automated enforcement in languages other than English, particularly where accurate interpretation requires an understanding of cultural and linguistic nuances and context.
- How should social media companies protect accounts from over-enforcement, especially journalists and political commentators who rely on social media accounts for their reporting and commentary.
- Whether automated enforcement without human review provides sufficient procedural safeguards when removing political speech and imposing account penalties.
- Good industry practices in transparency reporting on account enforcement decisions and related appeals.
In its decisions, the Board can issue policy recommendations to Meta. While recommendations are not binding, Meta must respond to them within 60 days. As such, the Board welcomes public comments proposing recommendations that are relevant to this case.
