Translate page with Google

Pulitzer Center Update November 18, 2022

Universities and Students Investigate Use of AI Surveillance Tool in Response to Grantee’s Reporting

Demonstrators at UNC-Chapel Hill protest the Silent Sam statue in August 2018.

How one company’s AI surveillance tool helps colleges nationwide monitor protests

Demonstrators at UNC-Chapel Hill protest the Silent Sam statue in August 2018. Image by Ari Sen. United States.

Student Journalists Take Matters of School Surveillance Into Their Own Hands

How do campuses protect and save the lives of students at risk of harming themselves or others? Software that uses artificial intelligence to scan students’ social media activities has been touted by companies and school officials as a promising way. 

What the story uncovered

Pulitzer Center AI Accountability Fellow Ari Sen and UC Berkeley Investigative Reporting Program journalist Derêka Bennett revealed a lesser-known use of an AI tool known as Social Sentinel: surveilling campus protests.

In a groundbreaking investigation for The Dallas Morning News, Sen and Bennett discovered that campus police across the U.S. used a technology called Social Sentinel (recently renamed to Navigate360 Detect) to track protests over a Confederate statue, student criticism of school officials over an alleged mishandling of a rape complaint, and even protests against a visiting U.S. senator at a town hall.

The reporting discovered that the company actively marketed the tool to university officials as an inexpensive solution to “mitigate” and “forestall” student protests while publicly maintaining that the service was not a surveillance tool. 

At least 37 colleges and universities across the United States—including Sen’s alma mater, the University of North Carolina (UNC)—were found to have used the tool since 2015. The Dallas Morning News published nearly 3,000 documents from the investigation, covering more than 56,000 pages from these schools. 

What happened after the story ran

Since publication, at least 11 universities and student newsrooms have used the reporting and public documents to begin inquiries into their schools’ use of Social Sentinel, and at least one university has dropped its contract with the company that sells the tool. The ripple effects of this investigation include more student reporting reacting to the broader implications of school surveillance on campuses across the country.

Seven universities in North Carolina were named on the list. Days after the story was released, the University of North Carolina (UNC) announced that it would terminate its relationship with Social Sentinel/Navigate360 Detect. In response to the story’s revelations, Graig Meyer, a state legislator for the area covering UNC, launched an inquiry into North Carolina colleges’ use of the surveillance technology.

Duke University students also probed into its use of Social Sentinel. Journalists at the university newspaper The Chronicle reported on the school’s questionable use of the tool; officials denied tracking student activity. Wake Forest University also confirmed that it had been monitoring student social media for campus safety since 2017.

Many universities and student newspapers have followed suit, including UCLA, UC Davis, University of Texas at Dallas, Arizona State University, University of North Texas, East Carolina University, and Virginia Commonwealth University. They published their schools’ contracts, made publicly available by Sen’s investigation, and many newsrooms are conducting investigations of their own.

The University of Connecticut’s student newspaper, The Daily Campus, obtained the police record for an April 2022 incident in which campus police received a Social Sentinel alert and opened a case because of it. It also published a list of terms that the university police tracked on social media, including former university president Susan Herbst and prominent locations on campus. The editorial board wrote a scathing op-ed condemning the university’s use of the tool, titled “UConn Expenditures on Student Surveillance Set a Bad Precedent.” Journalist Nathan Henault urged UConn to end its contract.

This story has also made its way into The Harvard Crimson—while Harvard University was not named in the reporting, its Cambridge, Massachusetts, neighbor MIT was—with an op-ed from The Crimson’s editorial board admonishing the use of AI surveillance on campuses around the country: “There’s a fine line between watching and watching out for,” the op-ed begins.

Why does this matter?

It’s inspiring to see students across the U.S. taking matters of school surveillance into their own hands. Many young journalists have quickly responded to the facts revealed through Sen and Bennett’s reporting in order to hold school officials and campus leaders accountable for overreaches in privacy through the misuse of surveillance tools that have not been proven to improve student safety and well-being on campus. 

“Mental health is a clear and pressing concern on college campuses; it’s one that demands administrative attention. But Big Brother-esque surveillance is hardly the solution administrators should turn to,” The Crimson’s editorial board urged.

“It’s really heartening to see the way students, and particularly student journalists, have responded to the story,” Sen said. “I hope it will encourage more of them to pursue investigative reporting into their schools and community leaders.”

Sen is a member of the first cohort of the Pulitzer Center’s AI Accountability Network, which seeks to expand and diversify the field of algorithmic accountability reporting. One of the pillars of the network is the radical sharing of methods and data so other journalists can reproduce and expand the investigations. 

We encourage journalists to continue pursuing their own stories—and we think there are ways for students and local media outlets to examine the misuse of artificial intelligence further. Please reach out to the Pulitzer Center if you would like to dig further into how college campuses are using AI to surveil student protest activity across the country.


Logo: The AI Accountability Network


AI Accountability Network

AI Accountability Network


teal halftone illustration of two children, one holding up a teddy bear


Children and Youth

Children and Youth
Criminal Justice


Criminal Justice

Criminal Justice
an orange halftone illustration of a hand underneath a drone


AI Accountability

AI Accountability