The State of Online Comments 2026.

1. The audience is moving to platforms publishers don't own

In the United States, 54% now access news via social media and video networks, a sharp rise, while news websites and apps reach 48% (Reuters Institute, Digital News Report 2025). Globally, weekly news reach runs through Facebook (36%), YouTube (30%), Instagram and WhatsApp (19% each) and TikTok (16%). For under-25s, social platforms are the main source of news.

The strategic problem is ownership. Engagement that happens on a platform belongs to the platform: the audience, the data and the reach are rented, and the algorithm can cut them off overnight. A comment, debate or community on your own domain is the opposite, it builds a first-party relationship you control. See first-party data and the complete guide to comment systems.

2. Trust is flat and avoidance is high, so quality is the battleground

Trust in news sits at 40% globally and has been stable for three years, while 40% say they sometimes or often avoid the news and 58% worry about telling real from fake online (Reuters Institute, Digital News Report 2025). A toxic, unmoderated comment space makes all three worse; a civil, well-moderated one does the opposite.

This is why moderation is no longer a cost centre but a quality lever. Across the Logora platform, hybrid AI plus human moderation auto-handles around 85% of incoming content before a human sees it, keeping spaces civil at scale. See AI moderation, toxicity detection and human-in-the-loop moderation.

3. Reader revenue depends on an owned relationship

Only 18% pay for online news across 20 countries, and payment is highly uneven (Norway 42%, US 20%, Germany 13%, UK 10%) (Reuters Institute, Digital News Report 2025). Converting and retaining that minority depends on a direct, identified relationship, exactly what anonymous or platform-bound engagement cannot provide.

The Logora deployments show the mechanism at work. At Milenio, after replacing Facebook Comments with a first-party comment system behind its own sign-in, daily comments grew 150% in the first year, and 10 to 11% of new daily registrations now come from the comment widget. At Der Spiegel, 21% of subscribers cite the debates as a reason they keep their subscription. Read the Milenio and Der Spiegel cases, and the guide on comments as a registration channel.

4. What the platform data shows works

Across 23 press groups in 12 countries and 50M+ contributions moderated since 2019, the pattern is consistent. Participation grows and pays off when readers sign in with their own account, when the space stays civil through hybrid moderation, and when the format rewards good contributions through debates, reputation and reasons to come back. A flat, anonymous comment box does not produce the same registrations or retention. The formats that do are comments, structured debates and consultations behind SSO and gamification.

5. AI changes both discovery and moderation

AI is entering the news loop from both ends. 7% use AI chatbots for news weekly (15% of under-25s) and 27% are interested in AI-summarised articles (Reuters Institute, Digital News Report 2025). On the supply side, AI is what makes real-time moderation viable at platform scale. The publishers who structure their conversation data well, with clear, first-party, well-moderated discussion, are also the ones whose expertise is best positioned to be surfaced and cited as AI mediates more discovery.

The takeaway

In 2026, the comment section is not a legacy feature. Done properly, with first-party sign-in, real moderation and a place inside the product, it is the most direct and measurable relationship a publisher or a brand owns with its audience. Platform reach, algorithms and consent rules can all change overnight. The conversation that lives on your own site does not. Own it, or keep renting your audience from someone else.

Sources and method

This report is a synthesis. Industry figures are drawn from the Reuters Institute Digital News Report 2025 (Reuters Institute for the Study of Journalism, University of Oxford), with each figure attributed inline. Platform figures are Logora's own data across deployments (Der Spiegel and Milenio client results; 50M+ contributions moderated since 2019; ~85% automated moderation), referenced on /clients and status.logora.fr. No figure is estimated or invented; where a number is Logora's own, it is labelled as such.