MeMo:KI study on the acceptance of artificial intelligence
People view use of AI in journalistic newsrooms critically
Bochum and Düsseldorf, 20 January 2021
AI systems can help journalistic work not only in distributing news, but also in supporting research activities in terms of content and even in writing articles autonomously. But what do people in Germany think about such developments?
The latest research findings of the MeMo:KI research project show not only that people detect little improvement in journalistic quality, but also that many favor strongly regulating AI in journalism.
Despite the critical verdict on the use of AI in journalistic work, those surveyed believe that some journalistic tasks can certainly be performed better by using AI than human journalists. Editorial departments should therefore only use AI technologies in a well-founded and transparent manner.
The Opinion Monitor Artificial Intelligence [MeMo:KI] has been investigating people’s attitudes to AI issues every two weeks since May 2020. How intensively do people engage with this technology? Where do they see areas of application, and what do they fear? The last survey focused on the question: How do people perceive the influence of artificial intelligence on the future of journalism? The findings show that only about one in five agree with the use of AI in the media. This is quite a low figure, especially when considering, for example, that 72% approve of the use of AI in industrial production.
The overall feeling among the respondents is that artificial intelligence will barely improve the quality of journalistic products. For example, only a fraction believe that AI will have positive effects: only 17% expect an increase in objectivity, while only 10% think that the media will become more credible. Nonetheless, respondents in the Opinion Monitor believe that AI can certainly perform some journalistic tasks better than human editors, with a high value being attested in particular to AI for technical tasks such as “manipulation detection” and so-called “fact-checking”. However, very few believe that AI can write better texts than humans (12%).
It is obviously important to the respondents that the use of AI is clearly visible to readers, with 82% favoring the mandatory labeling of AI in news production. However, only 17% call for a complete ban on AI in journalism.
“In summary, we can thus reveal quite interesting contradictions”, explains project member Kimon Kieslich. “While people definitely perceive the advantages in performance of AI systems, this does not help raise their expectations of quality. Almost the opposite: the use of AI in journalism is viewed quite critically. Editorial departments should therefore only use AI systems in a reflective and conscious manner”.
About the project:
The Opinion Monitor Artificial Intelligence [MeMo:KI] is conducting a survey every two weeks to record developments in how people view AI, with the results being published every month on the project homepage of MeMo:KI. The Opinion Monitor Artificial Intelligence is a joint project of the Center for Advanced Internet Studies (CAIS) in Bochum and the University of Düsseldorf, and is funded by the Ministry of Culture and Science of the State of North Rhine-Westphalia. Besides the regular survey of 1000 randomly selected people, there is also a semi-automatic evaluation of how the German media covers the issue. The results of both surveys are freely available to the public. “We prepare the results of our research in an online format with interactive graphics and tables so that all interested people can see for themselves how society and the media deal with the issue of artificial intelligence”, says Marcinkowski.
Text: Dr. Victoria Meinschäfer (University of Düsseldorf)
To the University of Düsseldorf’s press release