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 Live Captioning and Transcription for Public Hearings

What it is for

Inform
  • MOME uses Zoom's auto-captioning and transcription to make public rulemaking hearings and webinars accessible, providing real-time captions and post-meeting transcripts for the public.
Accessibility
  • Auto-captioning supports accessibility for hearing-impaired participants. MOME also provides American Sign Language and human-typed CART services as a reasonable accommodation upon request.
Run by
Mayor's Office
Where
No fixed location
Kept
Retained unspecified — determined by MOME's publication decisions and Zoom's platform policy
Shared with
Accountable organization, Vendor

What it is for

InformAccessibility

The Mayor's Office of Media and Entertainment uses Zoom's AI-powered auto-captioning and transcription features during public rulemaking hearings and webinars. The system converts spoken audio into real-time captions and post-meeting transcripts. When a transcript is published, a human reviews and corrects it before release. Attendees' voices are processed by Zoom's AI, though Zoom states it does not use meeting audio to train its models.

What it collects and what happens to it

Data taken in

Biometric
Identifiable data
  • Speech audio from participants at MOME's rulemaking hearings and agency webinars. Voices are biological signals that may be identifiable to specific speakers.

Processing

Speech & Audio
  • Zoom's AI uses speech-to-text processing to convert spoken audio from meeting participants into captions (real-time) and transcripts (post-meeting). This is the core AI technique described in the register.

What it does

Sensing (Perceptive AI)
Autonomous
  • Zoom's AI automatically converts live spoken audio into structured text (captions and transcripts) without human intervention during the meeting. The system is described as fully autonomous in the register.
Creating (Generative AI)
Human decides
  • The system generates written transcripts from spoken audio. When MOME chooses to publish a transcript, a human reviews and corrects the AI-generated text before release.

Outputs

Generated content
Identifiable data
  • Text transcripts and live captions generated from participants' speech. These may contain identifiable statements attributed to named or recognizable individuals speaking at public hearings. Published transcripts are reviewed and corrected by a human before release.

Run by

Mayor's Office
  • MOME deploys Zoom's AI captioning and transcription features for public rulemaking hearings and agency webinars held by the Mayor's Office.

NYC AI Use Cases Register — Zoom (MOME)

Built by

Zoom Video Communications
  • Zoom is the off-the-shelf vendor providing the virtual meeting platform, including the AI-powered auto-captioning and auto-transcription features. The application receives regular updates from the vendor.

NYC AI Use Cases Register — Zoom (MOME)

Kept for

Retained unspecified — determined by MOME's publication decisions and Zoom's platform policy
  • The register does not specify a retention period. Transcripts published by MOME are retained as public records. Captions are ephemeral. Zoom's own data retention practices apply to meeting data held on its platform.
  • Duration: unspecified — determined by MOME's publication decisions and Zoom's platform policy

Shared with

Available to the accountable organization
  • MOME has access to the transcripts generated by Zoom's AI. A human at MOME reviews and corrects any transcript before publication.
Available to vendor
  • Zoom processes audio and generates transcripts as part of its platform. However, Zoom states it does not use customer audio, video, or other communications content to train Zoom's or third-party AI models.

Stored

Stored on 3rd Party Cloud
  • Zoom is a cloud-based SaaS platform. Meeting audio, captions, and transcript data are processed and stored on Zoom's third-party cloud infrastructure.
  • Duration: unspecified — subject to Zoom's platform data retention policy
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 TransparencyParticipants are entitled to know that Zoom's AI is generating automatic captions and transcripts during MOME public hearings and webinars. Zoom's data practices are described in Zoom's privacy policy and terms of service at https://www.zoom.com/. MOME discloses its AI use in the NYC AI Use Cases Register.
  • Right to Be Informed of AI UseMembers of the public attending MOME hearings or webinars have the right to know that AI is in use for captioning and transcription. MOME's use of Zoom AI features is disclosed in the NYC AI Use Cases Register. Participants should be informed at the start of any session where AI captioning is active.
  • Right to Correct Your DataIf a published transcript contains errors in the transcription of a participant's speech, individuals may contact MOME to request correction. A human reviews all transcripts before MOME publishes them, but errors may still occur. Contact MOME through the Mayor's Office of Media and Entertainment.

Risks and safeguards

  • Psychological harmInaccurate auto-transcription of speech at public hearings could misrepresent participants' statements, causing distress or reputational concern.Safeguard: when MOME publishes a transcript, a human reviews and corrects it before release. MOME also provides human-typed CART as a reasonable accommodation upon request.
  • Reputational harmAI transcription errors could lead to published text that misquotes or misidentifies speakers at public proceedings.Safeguard: human review and correction of any transcript MOME chooses to publish. Captions are ephemeral and not stored beyond the meeting unless a transcript is generated.