How Aftonbladet uses AI

Aftonbladet, the Swedish newspaper within Schibsted, launched the AI chatbot "Hej Aftonbladet" in a joint project with Norwegian sister title VG and Schibsted Media AI Labs. The Swedish version gives readers quick answers to questions about news, sport and entertainment, drawing exclusively from Aftonbladet's own journalism. Martin Schori, Aftonbladet's Head of AI and Innovation, credits the design with avoiding the hallucinations that can affect other chatbots.

The News Machines newsletter reported in February 2026, after an interview with Schori, that the chatbot draws on three years of Aftonbladet journalism through retrieval augmented generation on GPT-4o, handles around 50,000 questions a day and had answered 7 million questions in nine months. Schori also said it launched without clear KPIs and that Aftonbladet has no data on conversion from chatbot logins to paying subscribers.

Aftonbladet's AI Hub, an eight person team of journalists, developers and a UX designer, built "Valkompisen", a chatbot that answered readers' questions about the June 2024 EU elections. Readers had to log in to keep chatting. The Online News Association reports that the bot answered more than 150,000 questions, 60 percent of them written by users themselves rather than picked from suggestions, and took 18,000 questions on its first day. The News Machines newsletter describes Aftonbladet's election chatbots as specialized versions of its general chatbot, which answers from Aftonbladet's own journalism through retrieval augmented generation.

Aftonbladet, a Swedish newspaper, built AI generated article summaries into its content management system. They appear below the article lead as expandable text. The Online News Association reports a click through rate of 43% on the expandable summaries (53% among readers aged 19 to 36), says elsewhere in the same case study that 40% of readers who saw the summaries expanded them, and reports that readers who used them spent more time on the full article. The ONA also reports that AI generated headlines did not perform as well as hoped and that tools inside the CMS worked better than tools that needed a second platform.