The next speaker in this session at the 2026 International Communication Association conference in Cape Town is Michael Reiss, whose interest is in the impact of generative AI on news consumption. Generative AI chatbots are now used in a wide range of informational contexts, including for exploring news topics; AI functionality is now also deeply embedded into search engines and other contexts.
At the same time, many people are actively avoiding the news, due to low levels of trust in the media, news overload, and other factors; will the growing role of generative AI also address news avoidance, then, as its outputs provide access to the news at a hyperpersonalised level to AI users, in a conversational format and on any topics of interest?
This might result in highly selective news experiences, limited only to topics of interest; it might present news in a more positive or constructive manner than generic mainstream news coverage; it might explain news topics more clearly and at the level of the user’s comprehension than standard news articles; there may be spillover effects from other AI uses to the exploration of news topics; and there may even be greater trust in the outputs of AI chatbots than in general news media.
However, AI outputs are potentially incorrect and unreliable; their use further undermines news outlets’ business models; and AI systems are unlikely to be used by audiences with already low digital literacies.
These potential developments were examined here through an online survey of some 1,500 participants in Germany in June 2024, just ahead of the EU elections. The survey measured news consumption, intentional news avoidance, AI use for news and political information in general, as a news substitute, and for the EU election. AI uses are still quite low (63% never use it for news), and lower still for very specific purposes; low news consumption and high news avoidance have no significant effect on such uses, either.
Those who use AI for news actually use more journalistic news sources, in fact; this further widens the gap between news users and non-users. The need to actively prompt AI for news might serve as a further barrier to news access here. However, we are still at an early stage of generative AI use for any purposes, of course; these patterns may yet change substantially over the coming years, especially also as people’s approaches to engaging with AI evolve and diversify.
