A groundbreaking study from RIT uses AI to analyze media coverage of police and challenges common public perceptions

Rochester, New York – A groundbreaking study led by researchers from Rochester Institute of Technology (RIT) and Carnegie Mellon University is challenging assumptions about the way local news media cover policing in the United States. Using advanced artificial intelligence (AI) techniques, the team analyzed a decade’s worth of reporting from local news outlets and found that coverage of law enforcement has remained largely stable over time—contrary to public perception that media has become increasingly critical of the police.

The study, recently published in the Proceedings of the National Academy of Sciences (PNAS), examined over 250,000 news articles from 10 politically diverse cities across the country. By applying large language models and natural language processing methods, the researchers systematically tracked trends in reporting from 2013 to 2023.

Stable Reporting, Shifting Perceptions

One of the study’s most striking findings is that criticism of the police in local media has not increased significantly over the past decade. While temporary spikes in critical coverage followed major incidents, such as the murder of George Floyd in 2020, these surges did not translate into a long-term shift toward more negative reporting. In fact, since 2020, local news coverage supporting the idea that police are effective has slightly increased.

Read also: Roc’n Ritmo makes a highly anticipated return to downtown with an exciting new date and legendary music lineup

“The data told a different story than we expected,” said Ashique KhudaBukhsh, an assistant professor in RIT’s Golisano College of Computing and Information Sciences and co-author of the study. “Local journalism on policing has been stable, but how people read the news depends on their political preference. The polarization is less on the news writers and more about the readers.”

To assess public perception, the researchers conducted a survey before launching their AI-driven analysis. The results showed that a majority of respondents—regardless of their political affiliation—believed that news reporting on the police had become more critical over time. However, the empirical data did not support this belief, suggesting that factors beyond the actual news content, such as selective exposure or preconceived notions, may be shaping public opinion.

A Closer Look at Political Influence

Unlike national media, which has been widely documented as shaping coverage to align with audience expectations, local news reporting on police did not appear to vary significantly based on the political leanings of the city. The study gathered data from 209 news outlets across five Republican-leaning and five Democratic-leaning cities, ensuring a balanced comparison of coverage trends.

Read also: Monroe County Executive Adam Bello shares exciting news about the return of the 2025 mobile household waste and pharmaceutical collection program

KhudaBukhsh and his colleagues found no substantial evidence that local journalists tailor their police reporting to match the political inclinations of their audience. This challenges the narrative that media bias is driving polarization at the community level, at least in the realm of police coverage.

“Our goal was to look at a broad time horizon—10 years—instead of focusing on a single moment in history,” said KhudaBukhsh. “By doing so, we were able to capture long-term trends rather than reacting to short-term events.”

AI and Human Judgment Work Together

To build a more nuanced understanding of how police coverage is framed in local news, the researchers incorporated diverse human perspectives into their AI model. Around 500 individuals—Republicans, Independents, and Democrats—reviewed excerpts from news articles and categorized them based on whether they supported or contradicted two key statements: “The police protect citizens” and “The police are racist.”

This process helped fine-tune the AI system to ensure it fairly represented different viewpoints and avoided political bias in its analysis.

“When building AI systems, it’s critical to include a diversity of viewpoints,” KhudaBukhsh explained. “If we only allow one political perspective to annotate the data, we risk creating an unbalanced evaluation with bias. Every data point in our annotation study was reviewed by a Republican, a Democrat, and an Independent.”

AI’s Expanding Role in Media Research

The collaboration between RIT and Carnegie Mellon highlights the growing role of artificial intelligence in studying large-scale media trends. While previous research has focused primarily on national news networks like MSNBC, Fox, and CNN, this study brings new insight into local journalism’s handling of policing topics.

“This is the best time to do this kind of research because the AI landscape has changed and we have the tools and data to ask deeper questions,” said Sujan Dutta, a computing and information sciences Ph.D. student at RIT and co-author of the study. “The biggest challenge was introducing the notion of political perspectives into AI and making a very comprehensive and balanced design that is representative of different viewpoints.”

Implications for Journalism and Society

The findings provide a data-driven perspective on media’s role in shaping public perceptions of law enforcement. By revealing that local journalism has remained relatively consistent in its police coverage, the study challenges narratives that claim media bias is fueling societal divisions.

“Media informs a great deal about our current state of the world,” said KhudaBukhsh. “By better understanding how sources cover the news and people consume it, we can work toward bridging the polarizing gap that we have today.”

The researchers hope their work will encourage further studies on media narratives and public perception, particularly as AI continues to evolve as a tool for large-scale analysis. Their approach demonstrates how technology can help answer complex societal questions by analyzing vast amounts of data with greater accuracy and depth.

In a time of increasing political division and debate over the role of media, studies like this provide a crucial reminder that perception does not always align with reality. With AI-powered tools and a commitment to balanced research, RIT and Carnegie Mellon are helping to uncover the truth behind the headlines.