In an age of digital communication and mass information sharing, online platforms play a powerful role in shaping public perception. Even seemingly neutral content can become a trigger for hate speech when published in environments that lack effective moderation. A recent case monitored by our team illustrates just how easily anti-Roma sentiment can erupt—not within the content itself, but in the user reactions beneath it.

The case involves a journalistic Facebook post reporting on the demolition of homes in the "Kumluk" neighborhood of Burgas, inhabited predominantly by Roma families. The article, framed as an objective piece, features interviews with Roma residents whose houses were destroyed by local authorities on the grounds that the buildings were illegal.

The text itself does not contain direct hate speech. There are no explicit insults, threats, or incitement to violence from the journalist. However, the title of the post—“Roma in Kumluk furious over their demolished illegal houses”—raises concerns about how language can shape reader attitudes. The use of emotionally charged phrases like “furious Roma” and “illegal houses” can reinforce negative stereotypes and invite judgment, even if unintentionally.

What followed in the comments section was a flood of dehumanizing, racist, and hate-filled responses. Among the most frequently encountered were:

• “Well done! Demolish them all!”
• “Pay your electricity, water, and taxes before you start shouting!”
• “Stop having children—our taxes are feeding you!”
• “You ate kebapcheta during the elections—now eat People’s salami!”

This pattern is not new. Our monitoring efforts repeatedly show that when a post contains the word “Roma” (or its derogatory equivalent “Gypsy”), the comments section often becomes a breeding ground for hate speech — regardless of the actual tone or intent of the post.

While social media platforms have made some progress in tackling hate speech in primary content, comment sections remain largely unmoderated and unregulated. Algorithms designed to detect harmful language rarely extend their reach to user replies, allowing bigotry to thrive just beneath the surface.

This case underscores the urgent need for more advanced moderation tools and greater platform accountability. Monitoring efforts must evolve to track not only published content but also the reactions it generates. We call on social media platforms, policymakers, and civil society actors to prioritize the development of technologies and standards that address hate speech in all its forms—especially those hidden in plain sight.

At TAAO, we remain committed to documenting and challenging anti-Roma hate speech in both its overt and subtle forms. This work is not just about defending the Roma community—it’s about protecting the integrity of our digital public spaces.