Generative AI and deceptive news consumption
Luca Sandrini & Róbert Somogyi
Highlights
• Users consume quality news and deceptive content (clickbait or fake news).
• Consumption of news can be direct (on news outlets) or indirect (on a GenAI portal).
• Early improvement of AI harms consumers and news outlets, benefits deceptive outlets.
• Incentives to produce news are lowered by early improvements in AI.
• Improvements in sufficiently mature AI benefit consumers but not news outlets.
Abstract
In this paper, we analyze the effects of advancements in generative Artificial Intelligence (GenAI) on the news media market. We model a representative consumer who allocates their time between reading news and deceptive articles. We find that GenAI may induce consumers to inefficiently reallocate their time and increase the consumption of the lower value good, i.e. deceptive content (clickbait articles or fake news). Therefore, early-stage GenAI distorts the incentives of consumers and reduces their welfare. After GenAI technology reaches a certain threshold, however, consumers start benefiting from its advancements. Finally, we find that the negative effects of early-stage GenAI are exacerbated as they induce a lower level of investment in news production.
