New York City, NY

A Billion-Dollar Fix That Doesn’t Fix the Problem

New York City’s subway system is facing a real crisis. Fare evasion has exploded, rider confidence has eroded, and the cost is borne by working New Yorkers who still play by the rules. The Metropolitan Transportation Authority now claims artificial intelligence is the answer. But beneath the glossy promises and vendor pitches lies a troubling reality: surveillance without accountability.

Below is the full analysis as originally written, preserved in its entirety.

New York Has a Real Problem. But This AI “Solution” Isn’t One

If you walk into many NYC subway stations today, you see people jumping over gates, slipping under turnstiles, or sailing through without paying. That’s unacceptable. Every fare evader shifts the burden to honest taxpayers who fund the system. The public wants accountability and enforcement, not gimmicks. Yet the new MTA system is being sold as a high-tech fix when what it really is amounts to surveillance theater.

Ramsey Khalifeh’s Gothamist article explains the superficial features of the pilot fare gates. But it does not answer the hard questions commuters deserve to know:

• How does this AI actually work?
• Does it stop fare evasion or just record it?
• Is the MTA building a database of images and profiles on riders?
• What enforcement follows after someone is flagged?
• Has fare evasion dropped since officers were removed from stations?

We need answers.

So What Is This Technology, Really?

According to the Gothamist piece, the new gates:

• Have cameras that record a short 5-second video clip when someone allegedly evades the fare.
• Use AI software to generate a written description of the person—height, gender, clothing, appearance—and send that to the MTA.
• Are part of a pilot at a handful of stations, with expansion planned to dozens more.

But this description raises real concerns:

1. Is the MTA building a profiling database?
   The article says nothing about retention policies.
   Does every video clip stay on file?
   Is the AI creating searchable biometric or descriptive profiles that get stored centrally?
   If the technology can describe you, it could also identify you. That’s a privacy red flag.
2. Who gets access to the footage?
   Is this only MTA employees?
   NYPD? State police? Private contractors?
   If a company builds this system, does it retain or share rider data?
3. What happens after someone is flagged?
   The article doesn’t clarify whether alerted fare evaders are stopped, fined, banned, or prosecuted.
   If no human enforcement follows, the technology is just a camera posting.

The public deserves clarity on these points before a $1.1 billion contract is put in place.

Modern Tech Won’t Fix What Enforcement Fixes

Here’s a blunt truth: fare evasion is a behavior problem, not a tech problem. Technicians can build clever turnstiles, but if there is no meaningful enforcement—no consequence for skipping the fare—the problem doesn’t get better; it just gets more high-tech.

Under the old system, when fare enforcement agents and police were present at busy stations, evasion dropped. The presence of a uniform served as deterrence. Honest riders were more likely to pay when they saw consequences.

When the MTA removed police from stations in 2021, citing concerns about equity and public safety, fare evasion rose sharply. Transit unions and independent analysts have documented this:

Time Period
Early 2010s with strong enforcement: approximately 5–8% of riders
Post-2018 removal of enforcement focus: approximately 15–25% or higher

These are estimates because the MTA does not publish transparent, independently audited numbers. But independent reporting and MTA internal presentations have consistently shown a significant rise after enforcement staff were pulled from stations.

Even with new gates, if the system only alerts but does not act, riders will exploit it.

AI Is Not a Substitute for Accountability

Reading the Gothamist article, a casual commuter might think the MTA finally has a high-tech miracle. In reality:

• The AI records behavior but does not stop it.
• No mention is made of issuing fines, summonses, or banning chronic offenders.
• No transparency is given on data storage, retention, or legal protections.
• The system could lead to a vast new database of subway rider imagery without meaningful privacy safeguards.

This is not accountability. It’s surveillance creep.

Riders across America are rightly skeptical of systems that collect and catalogue citizen behavior without clear purpose, rules, or oversight. The MTA should learn from those mistakes, not replicate them.

What Should Happen Instead

The MTA could do better by:

1. Publishing clear data on fare evasion trends both before and after enforcement changes.
2. Defining what happens after someone is flagged by AI. Is it a deterrent? A ticket? A court summons?
3. Establishing privacy safeguards for video, descriptions, and rider profiles.
4. Restoring meaningful human enforcement during peak hours to actually deter fare skips.
5. Measuring outcomes transparently so taxpayers know whether tech investment reduced evasion.

A government agency should be accountable to riders, not to vendors pushing expensive hardware.

A System That Watches but Doesn’t Protect

AI can be a tool. But when used without clear goals, meaningful enforcement, or safeguards, it becomes a cost nobody needs and a risk riders shouldn’t bear. The MTA must answer the tough questions before spending over a billion dollars on a system that watches us, but may not protect the transit system we all pay for.

References

• Gothamist — Modern NYC subway gates tested by the MTA use AI tech to track fare evaders
• MTA Budget Hearings and Internal Presentations
• Transit Workers Union Statements on Fare Evasion Trends
• New York State Comptroller Reports on MTA Revenue Loss

‍

Photo by Liliana Drew