AI News Monitor

The Metropolitan Transportation Authority is exploring the use of artificial intelligence to detect potential criminal activity in the subway before it occurs, part of a broader effort to enhance safety in the nation’s largest transit system.

The AI-powered surveillance system, still in development, would analyze live camera feeds to flag unusual or dangerous behavior—such as erratic movements or physical altercations—and alert transit police in real time. Officials say the goal is to intervene before crimes happen, a strategy they call “predictive prevention.”

“If someone is acting out, irrational… it could potentially trigger an alert,” said Michael Kemper, the MTA’s chief security officer and a 33-year NYPD veteran, during a safety committee meeting this week. “Before waiting for something to happen, we can respond.”

The MTA has not disclosed which tech companies it is working with or exactly what behaviors the system will target. But a spokesperson confirmed that the technology will not use facial recognition, focusing instead on patterns of movement rather than individual identities.

The initiative comes amid fluctuating crime rates in the subway. While major offenses dropped by 7.7% in early 2024, assaults have risen, prompting Gov. Kathy Hochul to deploy 1,000 National Guard members and state police officers to patrol stations. The AI system, officials say, is meant to supplement—not replace—human enforcement.

New York is not the first city to test AI surveillance in public transit. Los Angeles and Chicago have experimented with similar systems for weapons detection and parking violations. But civil liberties advocates warn of potential biases in AI algorithms and the risk of over-policing certain communities.

For now, the MTA assures riders that privacy protections are a priority. About 40% of subway cameras are actively monitored; AI could expand coverage without additional personnel. Still, questions remain about how the system will distinguish between harmless behavior and genuine threats—and whether it will deliver on its promise of a safer commute.