arrow_back Start page

Blog

Notes on comment moderation, online hate, and keeping public discourse peaceful.

Two protesters holding a cardboard sign and a megaphone during a peaceful demonstration (photo: Pexels, https://www.pexels.com/photo/protesters-holding-cardboard-and-megaphone-8898655/)

September 7, 2026

A Picture Is Worth More Than a Thousand Hateful Words, or On Moderating Images

Paste a slur into an image instead of typing it into a text box, and it can slip past even commercial content filters – research shows this. Memes and screenshots are also a real vector for health and political disinformation. ZenFeed now transcribes image comments and moderates them exactly like text.

By Szymon Nieradka

Close-up of a smartphone screen with social media app icons – Instagram, Facebook, Messenger and YouTube – on a dark background (photo: Pexels, https://www.pexels.com/photo/close-up-shot-of-app-icons-5678243/)

September 5, 2026

Automatic AI Comment Moderation on YouTube β€” the Next Platform in ZenFeed

X, TikTok, YouTube or Telegram? We compared the four biggest platforms on hate speech and disinformation, the comment moderation tools they give creators and brands, and how they work with law enforcement. Verdict: YouTube is the next platform in ZenFeed.

By Szymon Nieradka

Glowing Facebook logo above a backlit laptop keyboard (photo: Pexels, https://www.pexels.com/photo/illuminated-keyboard-with-symbol-in-focus-33427651/)

September 1, 2026

The Brzoska vs. Meta Dispute Shows Something No Platform Will Fix For You

RafaΕ‚ Brzoska is fighting Meta over billion-euro fines for scam ads built on deepfakes – and it's a fight worth winning. But it also shows that even a billionaire with a legal team waits weeks for a platform to react. A small page admin waits exactly the same way.

By Szymon Nieradka

Team of employees discussing a project while standing over a colleague working at a computer (photo: Pexels, https://www.pexels.com/photo/employees-discussing-project-standing-over-programmer-working-at-a-computer-6804079/)

August 20, 2026

Will Moderating Hate Reduce Engagement? What Research Actually Says

Many brands, creators and organizations hesitate to moderate toxic comments for fear of losing engagement. Research on comment moderation suggests otherwise: well-designed moderation can reduce toxicity while preserving β€” and sometimes increasing β€” meaningful participation.

By Alejandra Misiolek

Blue street sign casting a sharp shadow on a bright yellow wall (photo: Pexels, https://www.pexels.com/photo/street-sign-with-shadow-on-yellow-wall-30974520/)

August 18, 2026

Toxic Content Keeps People Online: Why Social Media Platforms Have a Conflict of Interest

A field experiment by Beknazar-Yuzbashev and colleagues found that when toxic content was automatically hidden, users spent less time on social media and saw fewer ads. This suggests a structural conflict of interest: reducing online hate may protect users, but it can also reduce platform revenue.

By Alejandra Misiolek

Elegant woman smiling joyfully during a night out (photo: Pexels, https://www.pexels.com/photo/joyful-night-out-with-elegant-woman-37191979/)

August 16, 2026

Online Hate Is a Mental Health Issue: What Selena Gomez, Billie Eilish and Other Public Figures Teach Us About Digital Wellbeing

Online hate is not just "mean comments." From Selena Gomez to Billie Eilish and Kate Winslet, public figures are speaking about the mental health impact of toxic social media. Here's why safer comment sections matter β€” and how ZenFeed helps.

By Alejandra Misiolek

Illustration of a boy running away from hands and hostile speech bubbles emerging from a phone, symbolizing online hate

August 13, 2026

Hate Speech Among Children and Teenagers: Why 'Bad Comments' Can Have Real Psychological Consequences

Data from SWPS, WHO/Europe and the Dbam o MΓ³j Z@siΔ™g Foundation show that 15-24 year-olds are the age group most exposed to online hate. What the research says about the mental health toll, and what adults and technology can do.

By Alejandra Misiolek

Tennis player lunging for the ball during an outdoor match on a clay court (photo: Pexels, https://www.pexels.com/photo/action-packed-outdoor-tennis-match-on-clay-court-37112809/)

August 10, 2026

When a Lost Set Turns Into Death Threats: Online Abuse in Women's Tennis

WTA and ITF data show that most direct abuse against tennis players comes from angry gamblers, and the psychological cost is real. Why prevention, not just moderation, is becoming part of athlete safety.

By Alejandra Misiolek

Close-up of a smartphone screen showing the Facebook app icon (photo: Pexels, https://www.pexels.com/photo/close-up-shot-of-a-smartphone-8296051/)

August 5, 2026

Why Facebook Comment Sections Feel So Toxic β€” and What the Science Says We Can Do About It

Toxicity isn't just unpleasant, it's engaging β€” and that's exactly why comment sections spiral. A look at the research on why Facebook hate feels so contagious, and what actually reduces it.

By Alejandra Misiolek

ZenFeed Moderation panel showing four Facebook comments with their toxicity score and the action taken

August 4, 2026

How does autonomous comment moderation actually decide? A real example from the ZenFeed dashboard

Four comments on the same post, three different toxicity scores, three different decisions. A real (anonymised) screenshot from the Moderation panel shows when the system acts on its own and when it hands the call to a moderator.

By Szymon Nieradka

European Union flags waving outside the European Parliament building in Strasbourg (photo: Pexels, https://www.pexels.com/photo/european-union-flags-at-strasbourg-parliament-39049421/)

August 1, 2026

GDPR and Facebook Comment Moderation: What's Allowed and What Isn't

Does deleting comments violate GDPR? What data does an automated moderation tool actually process, and what should you require from a vendor?

By Szymon Nieradka

Group of employees collaborating at computers in an open-plan office (photo: Pexels, https://www.pexels.com/photo/people-working-in-the-office-5453847/)

July 31, 2026

Facebook Comment Moderation: Manual or Autonomous?

Manual moderation works fine until it doesn't scale. We compare both approaches and show when a hybrid model actually makes sense.

By Szymon Nieradka

Top-down view of a smartphone and a pair of eyeglasses on a desk (photo: Pexels, https://www.pexels.com/photo/top-view-of-a-smartphone-and-eyeglasses-8970672/)

July 30, 2026

How Hateful Comments Damage a Brand's or NGO's Reputation

Toxic comments under a post hurt a brand's reputation even when the brand didn't write them. Here's why that happens and what you can actually do about it.

By Szymon Nieradka