How can AI tools help newsrooms moderate comments?

For many newspapers, the comments section is a great space for readers to contribute diverse thoughts, experiences and perspectives, but it can also become a breeding ground for negativity with arguments and offensive language that drowns out valuable voices.

This was the challenge faced by Financial Times community editor Lucy Warwick-Ching. The FT has a dedicated moderation team, but with hundreds of articles published each week, they couldn’t keep up with the pace, and comments were closed on most of them. They also decided to close all comments sections on weekends to avoid toxic discussions spiraling out of control when no one was at their desks.

Hiring additional moderators was too expensive and building their own moderation system was too cumbersome. So Warwick-Ching's team used a pre-existing moderation tool from Utopia Analytics, which was then trained on 200,000 real reader comments to recognize polite, non-argumentative, and off-topic discussions.

AI tool can identify sexist and racist comments, despite complex language. Photo: Eidosmedia

When the tool first started, the FT still had to manually check every comment it approved. It took several months for the tool to make the right moderation decisions. It now catches most sexist and racist comments, despite the complex language FT readers use.

“It's not perfect and it's still a work in progress,” Warwick-Ching said after six months.

Still, the impact is significant. Previously, moderators spent most of their time filtering out negative comments. Now, AI does a lot of the heavy lifting, allowing moderators to focus on building communities.

Not only is the AI ​​tool effective in moderating comments, it also replaces humans in dealing with emotionally damaging comments. AI now absorbs much of the negativity that comes from those comments, protecting humans from the negativity.

Most importantly, Warwick-Ching says the introduction of AI tools doesn't mean any moderators are out of a job. "From the beginning when we introduced the moderation tool, we made it clear that this wasn't about reducing staff, but about creating a space for moderators to do their jobs."

So far, the AI ​​tool has yielded positive results. Readers have emailed to say they’ve noticed improved comment civility, and there haven’t been any significant complaints about comments being unfairly rejected.

Journalists also saw a significant increase in the benefits of comments, with valuable reader input enriching their stories. They also found that using AI allowed moderators to focus on what they do best: fostering a welcoming and informative space for people.

Hoai Phuong (according to Journalism)

Source: https://www.congluan.vn/cong-cu-ai-co-the-giup-cac-toa-soan-kiem-duyet-binh-luan-nhu-the-nao-post298847.html