A recent study published in Social Science Computer Review finds that South Korea’s conservative and progressive newspapers frame artificial intelligence issues differently, reflecting their distinct ideological orientations. Conservative outlets tend to emphasize the benefits and advancements associated with artificial intelligence, while progressive media more frequently highlight ethical concerns and risks. These different portrayals influence public understanding and have implications for policy discussions about artificial intelligence.
The study was conducted due to growing public and political interest in artificial intelligence technology. As artificial intelligence becomes increasingly prominent in our daily lives, understanding how the media present these issues is important. Media coverage shapes public opinion and influences how society approaches technology, as well as affecting government policy decisions.
Given South Korea’s global leadership in technological innovation and its politically polarized media landscape, researchers recognized an opportunity to explore how media partisanship influences the presentation of artificial intelligence news. The research team aimed to identify how conservative and progressive newspapers differ in their framing of artificial intelligence topics, hoping to offer insights into the wider societal implications of these differences.
To conduct her analysis, Mikyung Chang from Korea University examined articles published by two of South Korea’s most prominent newspapers, each representing a distinct political ideology. They selected Chosun Ilbo, representing conservative viewpoints, and Hankyoreh, known for progressive perspectives. Chang collected 684 articles published from November 30, 2022, to June 30, 2023, focusing on topics related to artificial intelligence, such as ChatGPT and generative artificial intelligence. The date was significant as it marked the public introduction of ChatGPT, sparking heightened interest and debate around artificial intelligence.
After eliminating articles unrelated or minimally related to artificial intelligence, Chang narrowed the sample down to 584 articles for detailed analysis. Two trained coders independently analyzed each article, categorizing them according to two main value frames: a “development” frame, which emphasizes positive aspects like economic growth, technological progress, and quality-of-life improvements, and a “crisis” frame, highlighting ethical concerns, risks, and potential negative impacts. Chang ensured accuracy through detailed coding criteria and verification of consistency between coders.
The analysis revealed clear differences in how South Korea’s conservative and progressive newspapers covered artificial intelligence. Conservative media, exemplified by Chosun Ilbo, predominantly used the “development” frame. They emphasized artificial intelligence’s positive aspects, highlighting how technological advancements can lead to economic growth, improved quality of life, and enhanced competitiveness.
Conservative newspapers often portrayed artificial intelligence as a promising opportunity, pointing out its potential to drive economic growth, enhance productivity, and improve healthcare and education. For instance, articles emphasized how generative artificial intelligence could significantly boost economic output, create new business models, and foster innovation beneficial to society as a whole.
In contrast, the progressive media, represented by Hankyoreh, were more likely to use the crisis frame. Progressive coverage frequently addressed the ethical, social, and legal concerns surrounding artificial intelligence. Issues such as privacy risks, job displacement, ethical dilemmas, and the dangers of misinformation and manipulation were common themes.
Articles often discussed the potential for artificial intelligence-generated fake news and privacy violations, along with job displacement resulting from automation and generative artificial intelligence. The progressive media also highlighted concerns over increasing dependence on artificial intelligence and called for societal preparation and appropriate regulations.
The difference in coverage approaches between conservative and progressive outlets reflects deeper societal and ideological priorities. Conservative media appear inclined toward maintaining public optimism, supporting the notion of technological advancement and economic prosperity. Progressive media, conversely, emphasize caution, responsibility, and the need to address potential ethical and social problems. This division in media framing indicates distinct goals for public debate—conservatives prioritize technological progress, while progressives focus on equitable development and the prevention of potential negative consequences.
Chang emphasized that such framing shapes how citizens perceive artificial intelligence, influencing whether they view these technologies as beneficial or harmful. By highlighting specific aspects, media frames can shape public opinion and influence how society and policymakers respond to technological innovations.
“This research provides a broad understanding of how media shape public opinion on new technologies and can inspire future studies to investigate whether partisan characteristics and patterns in AI reporting are present in other technologically advanced nations and how they influence public perceptions and behavioral intentions,” she concluded. “By exploring media framing of AI issues in various geopolitical contexts, such research could enhance our comprehension of global AI discourse and aid in devising strategies for addressing social and ethical issues related to AI. It can also contribute to the development of global regulatory guidelines and policies through effective science communication strategies.”
The new findings are in line with research conducted in the United States. A 2023 study found that liberal-leaning media outlets express more negative sentiments toward artificial intelligence compared to conservative-leaning outlets, primarily due to concerns over AI’s potential to reinforce societal biases. This partisan divide in AI coverage persisted even after accounting for factors like analytical depth and authenticity.
The new study, “Does the Media’s Partisanship Influence News Coverage on Artificial Intelligence Issues? Media Coverage Analysis on Artificial Intelligence Issues,” was published September 2, 2024.
