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 Closed Captioning for City Television

What it is for

Accessibility
  • The system generates closed captions for video content broadcast on NYC's television channels, making programming accessible to viewers who are deaf or hard of hearing.
Run by
Mayor's Office
Where
No fixed location
Kept
Retained Not specified in source materials
Shared with
Accountable organization, Vendor

What it is for

Accessibility

NYC Media uses Adobe Premiere Pro's AI-powered speech-to-text tools to help human editors generate closed captions for video content broadcast on New York City's television channels. The AI transcribes spoken words from video programs into caption drafts, which a human editor reviews before broadcast. The system is powered by Adobe Sensei machine learning and third-party speech recognition technologies.

What it collects and what happens to it

Data taken in

Biometric
Identifiable data
  • Spoken words captured in video programs — the voices of speakers appearing in content edited for broadcast on NYC television channels.

Processing

Speech & Audio
  • Adobe Premiere Pro's Speech to Text feature converts spoken words in video programs into text using Adobe Sensei machine learning and third-party speech recognition technologies.

What it does

Sensing (Perceptive AI)
Human decides
  • The AI listens to spoken audio in video programs and transcribes it into text caption drafts. A human editor reviews and approves all captions before broadcast.

Outputs

Generated content
Anonymized data
  • Closed caption text generated from spoken audio in video programs, reviewed by a human editor before being included in broadcast content on NYC television channels.

Run by

Mayor's Office
  • NYC Media deploys Adobe Premiere Pro with AI captioning tools to produce and broadcast video content on New York City's television stations.

NYC AI Use Cases — Adobe Premiere Pro

Built by

Adobe
  • Adobe provides Adobe Premiere Pro as an off-the-shelf commercial product with regular software updates, including the AI-powered Speech to Text feature built on Adobe Sensei machine learning and third-party technologies.

NYC AI Use Cases — Adobe Premiere Pro

Kept for

Retained Not specified in source materials
  • The register does not specify a retention period for AI-processed audio or generated caption data. The final broadcast video and its embedded captions are retained as part of normal media archive practices.
  • Duration: Not specified in source materials

Shared with

Available to the accountable organization
  • Caption output is accessible to NYC Media editors and staff within the Adobe Premiere Pro editing environment. Broadcast captions become part of the public television record.
Available to vendor
  • Adobe, as the vendor of Premiere Pro, may process audio data through its Speech to Text service. The extent of Adobe's data access and any subprocessors involved in the third-party speech technology are not detailed in the register entry.

Stored

Not stated by the NYC Office of Technology and Innovation.

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 Be Informed of AI UseViewers watching NYC television broadcasts should be aware that closed captions may be generated with AI assistance via Adobe Premiere Pro's Speech to Text feature. The register entry is publicly disclosed through the NYC AI Use Cases register.
  • Right to Algorithmic TransparencyAdobe's Speech to Text feature is powered by Adobe Sensei machine learning and third-party technologies. General information about the technology is available on Adobe's website at https://www.adobe.com/products/premiere.html.

Risks and safeguards

  • Reputational harmInaccurate AI-generated captions could misrepresent speakers' words in publicly broadcast content, potentially harming reputations.Safeguard: the system operates in a monitored mode where a human editor reviews all AI-generated captions before broadcast, maintaining editorial control over final content.