Interview on Artificial Intelligence and ChatGPT

For the Science Media Center, Andreas Jungherr classifies the role of artificial intelligence and ChatGPT in the creation and spread of disinformation

In the current discussion about the potential and risks of artificial intelligence (AI), the danger of disinformation is a recurring topic. In particular, there is the concern that disinformation can be generated more easily in greater quantity and better quality using language models. Other types of AI, such as image generators or software for creating deepfakes, could also play a role in the context of disinformation if increasingly realistic-looking images and videos can be easily created by AI.

But the impact of disinformation depends on many factors, not just the difficulty of creation. For example, the spread of the content or the target group's belief in its accuracy can also play a role. Therefore, the question arises as to the importance of individual factors and the role played by the creation facilitated by AI methods.

Since the topic is still quite new, especially with regard to the role of language models, there is not yet much completed research on the topic. But looking at previous research on the impact and spread of disinformation, researchers can already provide initial predictions and assess the state of the research and debate. Therefore, we asked researchers who conduct research on disinformation for their initial assessments on this topic.

Desinformation: Welche Rolle können Sprachmodelle und KI spielen?