Skip to content
This is NOT an official site of the City of New York. Click here for official AI reports.

NYC Office of Technology and Innovation

NYC - DTPR AI Register

AI-Assisted Two-Year Jail Population Forecast

What it is for

Planning & Decision-making
  • Projects the expected daily NYC jail population and associated uncertainty ranges over a two-year horizon to support high-level planning and budget decisions for the Department of Correction. Required by Terms and Conditions added to DOC's budget in 2024; reports are published publicly every six months.
Run by
Department of Correction (DOC)
Where
No fixed location
Kept
Retained Not specified in the register
Shared with
Accountable organization, Download

What it is for

Planning & Decision-making

This system uses 12 years of historical data to project the New York City jail population on a daily basis for the next two years, along with estimated uncertainty ranges for each day. It is used by the NYC Department of Correction for high-level planning and budget decisions. The system does not make decisions about individuals — it forecasts aggregate population levels to support resource allocation. Forecasts are published publicly every six months.

What it collects and what happens to it

Data taken in

Operational data
Anonymized data
  • Runtime inputs include the latest aggregated population levels, admission rates, and arrest rates. Also includes aggregated information for incarcerated individuals such as admission dates, historic changes in incarceration status (e.g., City Sentenced, Pretrial Detainee, State Ready), and length of stay.

Processing

Classification & Prediction
  • A time-series forecasting model built on 12 years of historical data. It generates daily population projections and prediction intervals (uncertainty ranges) over a two-year horizon via back-tested statistical modeling. The register describes the computation type as 'Forecasting'.

What it does

Deciding (Analytical AI)
Human decides
  • The system forecasts aggregate jail population levels and prediction intervals from historical trends. It is described as 'informative' in the register — outputs are advisory inputs to planning and budget decisions made by DOC leadership, not automated decisions.

Outputs

About a measurement
Anonymized data
  • The system outputs a daily forecast of the expected NYC jail population count and prediction intervals (uncertainty ranges) for each day over the next two years. Outputs are aggregate population figures — no individual-level data is produced.
A recommendation or prediction
Anonymized data
  • The forecast and its associated uncertainty intervals serve as advisory inputs for high-level planning and budget decisions within DOC. The autonomy is described as 'Informative' — no automated decisions are taken based on the output.

Run by

Department of Correction (DOC)
  • The NYC Department of Correction (DOC) operates this forecast system through its Operations Research department. DOC uses the tool for high-level planning, budget purposes, and publishes the reports publicly per Terms and Conditions added to DOC's budget in 2024.

NYC AI Use Case Register — NYC Jail Population Forecast

Built by

Not stated by the NYC Office of Technology and Innovation.

Kept for

Retained Not specified in the register
  • The register does not specify a formal retention period for the forecast outputs or the operational input data. Training data covers 12 years of historical records from internal sources.
  • Duration: Not specified in the register

Shared with

Available to the accountable organization
  • DOC's Operations Research department uses the forecasts internally for high-level planning and budget purposes.
Available to download
  • Forecast reports are published publicly every six months per the Terms and Conditions added to DOC's budget in 2024. The first report was released in January 2025.

Stored

Stored locally
  • Training data is from internal DOC sources. The register does not specify cloud or off-site storage arrangements; the system is operated internally by DOC Operations Research.
  • Duration: Not specified in the register
How to read the colours

Can it identify you?

Anonymized data
Data about people with the link to who is broken. Stripped of identifiers, blurred, aggregated, or noised so this system can’t reasonably tie a record back to an individual.
Pseudonymous data
Each person’s data is tied to a token (hash, ID, template) that lets this system recognise the same person across events, but the token itself doesn’t reveal a name. Reidentification is possible with extra information.
Identifiable data
The data either contains a direct identifier (name, address, account name, recognisable face or voice, plate number) or carries a token this system uses to look up legal identity during processing.

Who completes the loop?

Human decides
This mode suggests; a person decides what to do next. The AI is always advisory — a human is in the loop on every decision. Example: a triage tool ranks cases for a clinician who chooses which to see first.
Human executes
This mode decides; a person carries out the result. Example: an optimizer plans the day’s trash-collection routes, and drivers run them.
Autonomous
This mode decides and acts on its own. No person reviews each decision or carries out the resulting action.

Definitions from the DTPR standard. Amber is about your data, violet about who decides. The fuller the shape and the deeper the colour, the more identifying the data or the less a person is involved.

What you can do

Ask about this system

Questions go to the NYC Office of Technology and Innovation, not the vendor.

Your rights

  • Right to Algorithmic TransparencyThe forecast methodology and outputs are publicly disclosed through reports published every six months on the NYC Council Budget website. Members of the public can review the published forecast reports at https://council.nyc.gov/budget/.
  • Right to Be Informed of AI UseThe system operates on aggregated, non-individual data and does not make decisions about specific people. The existence of this forecasting tool and its methodology are publicly disclosed in reports published by the NYC Department of Correction every six months.

Risks and safeguards

  • Financial & business harmInaccurate forecasts could lead to misallocation of public resources — over- or under-provisioning of jail capacity, staffing, and budgets.Safeguard: The model undergoes rigorous back-testing to estimate prediction intervals and quantify uncertainty. Forecasts are published publicly every six months, enabling external scrutiny and accountability. The system is advisory only; final budget and planning decisions remain with DOC leadership.
  • Civil liberties harmPopulation forecasts for a jail system touch on the liberty of incarcerated persons; if used to expand incarceration capacity beyond genuine need, they could indirectly entrench over-incarceration.Safeguard: The system is explicitly described as informative and advisory. Forecasts are published publicly, allowing civil society and policymakers to scrutinize trends. No individual-level decisions are automated by this system.