Political Ideology Affects AI Coverage

As artificial intelligence (AI) becomes a part of our daily lives, from language translation to virtual assistants, it remains a topic of debate. Virginia Tech researchers are exploring which segments of society are more open to AI and which ones might be more resistant.

The researchers analyzed how media with different political leanings reacted to AI. The findings showed that liberal-leaning media tends to express more negative views on AI compared to conservative media. This suggests that liberal media is generally more opposed to AI than conservative media.

Gauging sentiment

To gauge the sentiment of partisan media toward artificial intelligence (AI), the researchers assembled a collection of articles about AI from various media outlets. They determined the partisan sentiment for each outlet by referencing the Media Bias Rating Chart from AllSides, a company that assesses the perceived political bias of online news content. The selection included both liberal-leaning outlets like The New York Times and The Washington Post, and more conservative-leaning outlets such as The Wall Street Journal and the New York Post.

The researchers then downloaded articles based on specific criteria, including the use of key terms like “algorithm” or “artificial intelligence,” within a date range from May 2019 to May 2021. The dataset comprised over 7,500 articles, and the researchers conducted an emotional tone analysis on each using an automated text analysis tool.

This tool measured the emotional tone of each article by calculating the difference between the percentage of positive and negative emotion words, standardized on a scale of 0 to 100 to determine the emotional tone.

Liberal concerns

The research suggests that this opposition is rooted in liberal concerns about AI amplifying social biases like racial, gender, and income disparities, as opposed to conservative media. The researchers also looked at how media sentiments about AI changed after the death of George Floyd.

“Since Floyd’s death ignited a national conversation about social biases in society, his death heightened social bias concerns in the media,” the researchers explain. “This, in turn, resulted in the media becoming even more negative towards AI in their storytelling.”

The researchers argue that the coverage in the media is often used to gauge public sentiment by politicians, which in turn influences the regulations and other guardrails introduced to manage the development of the technology.

“Media sentiment is a powerful driver of public opinion, and oftentimes policymakers look toward the media to predict public sentiment on contentious issues,” the researchers continue. “Perhaps the next step in our research is to see how social media conversations surrounding AI change as a function of the partisan differences we see in our paper.”

Facebooktwitterredditpinterestlinkedinmail