Ellen Yaeger
During the shootings at Umpqua Community College, Texas Southern University, and Northern Arizona University, the Georgetown University Police Department (GUPD) was put on high alert for any suspicious activity on campus. These tragedies and GUPD’s reaction highlights a reality of campus security—much of it occurs after-the-fact. What if GUPD could predict an active shooter event before it happened? Boston University is using new technology to do just that, creating both an interesting and troublesome precedent for university law enforcement.
Since 2012, 143 school shootings have devastated communities across the nation. Every university has procedures to deal with such events: Georgetown works with the Metropolitan Police Department to assess local threats. GUPD also collaborates with university mental health administrators to identify individuals at risk for violent behavior. Meanwhile, Boston University’s Police Department (BUPD) has begun to monitor social media networks to bolster it ability to identify potentially threatening individuals and prevent them from doing harm.
Last January, BUPD unveiled new software to monitor nine social media platforms, including Facebook, Twitter, Instagram, Flickr, and Yik Yak. This online tool tracks posts on these sites and flags those containing words associated with violence. In particular, the software reviews social media posts that have the location services feature enabled. Once a suspicious post is flagged, BUPD can pinpoint exactly where that post originated. The software then creates a “geofence” around that location, enabling the BUPD to conduct real-time analysis of all subsequent posts made in that area.
Moreover, BUPD’s software monitors the social media activity of anyone in the vicinity of BU’s campus—meaning BU students and faculty, as well as locals and individuals passing through the area can be tracked. BUPD’s software is not only useful for identifying potentially violent, unstable students, but also local criminals. For a school like Georgetown where crime frequently occurs in the surrounding neighborhoods by actors unaffiliated with the university, BUPD’s software seems like an appealing program.
But, the BUPD’s software raises at least as many problems as it solves. The software is certainly a powerful predictive tool, but it is also a possible invasion of students’ privacy. BU’s news and information website BU Today reported on the software days after the shooting in Roseburg, Ore. The initial reporting, however, fails to mention whether the software can access only public posts, or any type of post. Even if the software only accesses public posts now, what if later models are able to view posts intended for “Friends only” audiences or private messages?
The software is also likely to flag false positives. Many people use social media to rant; fortunately, not as many people shoot up schools. As a result, police might waste considerable time and money on false leads. In addition, students might have police records only because of their word choice on social media. It is also unclear whether BUPD developed the software itself or purchased it from an external vendor, which would clarify the possibility and ease of this technology appearing on other campuses.
Given these concerns, this Editorial Board finds BUPD’s software in need of more development before it can be truly effective in its aims—especially if other universities begin to adopt similar programs. Currently, the online tool appears to hold the potential to combat very serious threats that campuses across America face, but its implementation needs improvement. If Georgetown were to one day consider use of such software, we urge careful analysis and vetting of the product. BUPD’s software provides an intriguing functionality to combat issues of campus safety, but it also adds to the ongoing debate in our country about safety at the sacrifice of privacy.
