NYC Launches Pilot AI System to Accelerate Building Permit Reviews in Four Boroughs

NEW YORK — The New York City Department of Buildings launched a pilot program Monday to test an AI-assisted document review system across four borough offices, the latest in a series of technology investments aimed at reducing permit processing backlogs that have delayed construction projects — from small residential renovations to large commercial builds — for years. The tool, developed by a firm called PermitIQ in partnership with the department's technology unit, uses machine learning to pre-screen architectural drawings, engineering calculations, and code compliance documentation before they reach a human plan examiner.

The department processes approximately 280,000 permit applications per year, a volume that has grown substantially as the city's construction sector has expanded and as new code requirements have added documentation layers to even routine projects. The average time from application submission to permit issuance for a standard new building application currently stands at 38 days — up from 24 days in 2019 — and the backlog is longer for complex applications involving structural work, façade modifications, or projects in historically designated areas. Small contractors and homeowners making modest renovations are disproportionately affected, as they lack the resources to hire expediters who can navigate the system on their behalf.

PermitIQ's system works by ingesting submitted documents and flagging likely compliance issues, missing information, or drawing inconsistencies before a plan examiner opens the file. The model was trained on eight years of New York City permit applications — both approved and rejected — and the company says it achieves a deficiency detection rate of 91 percent on a held-out test set, meaning it correctly identifies nearly all the issues that would otherwise require a plan examiner to issue a comment letter and restart the applicant's clock. Applications that clear the pre-screening with no flags can be routed to an expedited examiner queue.

The pilot covers the Brooklyn, Queens, Bronx, and Staten Island borough offices; Manhattan, which handles the city's most complex commercial applications, will be assessed for inclusion after the initial pilot phase concludes. Department of Buildings Commissioner Rachel Teng said the goal is to demonstrate that AI pre-screening can compress review cycles without compromising code compliance or examiner judgment. "We are not replacing plan examiners," she said at a press briefing. "We are giving them cleaner files so they can spend their time on the decisions that require professional judgment, not on chasing down missing owner signatures or mismatched sheet numbers."

Architects and contractors who have long complained about the permit process reacted with cautious optimism. The American Institute of Architects New York chapter said it would monitor the pilot and called for the city to share performance data publicly on a quarterly basis. The Building Trades Employers' Association said that any technology that reduces project delays is good for the industry, while noting that processing speed is only one of many factors affecting construction timelines. Several small contractors expressed hope that the expedited queue would make a visible difference for the residential jobs that represent the bulk of their work.

Privacy and governance advocates have asked the city to publish the documentation on how the PermitIQ model was trained and validated, and to ensure that the pre-screening does not disproportionately flag applications from projects in certain neighborhoods or by applicants with less sophisticated document preparation. The Department of Buildings said it will conduct a bias audit of the pilot's outputs after 90 days and publish the results. The contract with PermitIQ is structured as a performance-based arrangement, with payments tied to measured reductions in average review time.

The pilot is expected to run through September, at which point the department will analyze outcomes and make a recommendation about full-scale deployment. If the results are positive, a citywide rollout could follow in early 2027. Commissioner Teng said the department is also evaluating AI tools for other high-volume functions, including elevator inspection scheduling and complaints triage, as part of a broader technology modernization plan that the administration has committed to funding over the next three fiscal years.