A look at New York City's slow attempts to increase transparency around its use of algorithms, which govern many policy areas, including policing
The Traffic Management Center for New York, located in Queens. Two tools used by the Department of Transportation were reviewed …
Context & Ripple Effects
New York City has been quietly wiring algorithms into its policy machinery for years: the Drive Smart pilot tracked the driving habits of 400 volunteers back in 2015, and by 2019 the Department of Transportation was running tools at the Queens Traffic Management Center that a Vox review put under scrutiny. The through-line is that these systems govern consequential decisions — traffic flow, policing support — with little public visibility into how they work.
What makes this story durable is what came after it: the city went on to pass [[a:840498|a law forcing firms using AI in hiring to notify candidates and submit to annual bias audits]], while its own deployments kept expanding — from MTA fare-evasion surveillance to Google's TrackInspect on subway tracks. Meanwhile the child-welfare risk-scoring tool launched in 2018 is now drawing racial-bias concerns, exactly the failure mode transparency advocates warned about.
First-order effects
- The Department of Transportation's two reviewed tools become test cases for whether city agencies can explain their own systems — the same agencies running them are the ones answering for them.
- Agencies deploying algorithms without disclosure, like the child-welfare scorer, face immediate credibility costs as bias concerns surface around systems launched years earlier.
Second-order effects
- Other city bodies with active deployments — the MTA's expanding station surveillance program chief among them — get pulled into the same transparency debate, since each new rollout raises the question the DOT review is asking.
- Vendors selling algorithmic tools to the city face pressure to build in explainability and audit access, because opaque systems are becoming a procurement liability.
Third-order effects
- If the pattern holds, New York's piecemeal reviews harden into a standing regime: mandatory notification and independent bias audits, already law for hiring AI, extending across city government as the default condition for deploying algorithms.
- Cities nationally would then treat algorithmic accountability as infrastructure governance rather than a one-off controversy — with procurement rules, not press scrutiny, doing the enforcement.
The trend: Municipal governments are moving from ad-hoc algorithm deployments toward formalized disclosure-and-audit regimes, with New York's slow, contested path setting the template other cities will copy or correct.