AI-Powered Live Captioning for Public Meetings

What it is for

Accessibility

Captions are generated to improve access for participants with hearing disabilities or language barriers, both during live Open Data Week events and in the YouTube recordings published afterward.

NYC AI Use Case Register — Zoom Automated Captions

Run by: Office of Technology and Innovation (OTI)
Where: No fixed location
Kept: Retained Indefinitely (no deletion schedule specified)
Shared with: Accountable organization, Download, Vendor

What it is for

Accessibility

This system uses Zoom's automated speech recognition to generate live closed captions and full transcripts during NYC Open Data Week public meetings. Captions appear in real time for meeting attendees and are later embedded in recordings uploaded to YouTube. The system is operated under a collaboration between the NYC Office of Data Analytics and the civic-tech non-profit BetaNYC, with the explicit goal of improving accessibility. The underlying speech recognition model is proprietary to Zoom.

What it collects and what happens to it

Data taken in

Biometric

Identifiable data

Live audio (voice) from Zoom meeting participants — speakers whose voices are captured by their microphones during Open Data Week events. Voice is a biometric signal tied to identifiable individuals.

NYC AI Use Case Register — Zoom Automated Captions

Processing

Speech & Audio

Zoom's proprietary automatic speech recognition converts live spoken audio from meeting participants into text captions. The training data behind the model is proprietary to Zoom and not disclosed.

NYC AI Use Case Register — Zoom Automated Captions

What it does

Sensing (Perceptive AI)

Autonomous

Zoom's speech recognition engine listens to live audio from meeting participants and autonomously converts it into structured captions and transcripts, with no human review of individual caption outputs before display.

NYC AI Use Case Register — Zoom Automated Captions

Outputs

Generated content

Identifiable data

Real-time closed captions displayed to meeting attendees, and a full transcript file embedded in the Zoom recording. Transcripts are uploaded to YouTube alongside the event video, where they become publicly accessible and identifiable to speakers.

NYC AI Use Case Register — Zoom Automated Captions

Run by

Office of Technology and Innovation (OTI)

The Office of Data Analytics, within the NYC Office of Technology and Innovation, deploys and uses Zoom Automated Captions for public Open Data Week meetings. BetaNYC, a civic-tech non-profit collaborator, owns and operates the Zoom account used for these events and has access to the transcription files.

NYC AI Use Case Register — Zoom Automated Captions

Built by

Zoom Video Communications / BetaNYC

Zoom provides the off-the-shelf automated captioning technology; its speech recognition training data is proprietary. BetaNYC provides professional services as the account owner and operator of the Zoom account, and handles uploading transcribed recordings to YouTube.

NYC AI Use Case Register — Zoom Automated Captions

Kept for

Retained Indefinitely (no deletion schedule specified)

Transcripts embedded in YouTube recordings are retained indefinitely as part of publicly published event archives. The register does not specify a defined deletion schedule.
Duration: Indefinitely (no deletion schedule specified)

NYC AI Use Case Register — Zoom Automated Captions

Shared with

Available to the accountable organization

The NYC Office of Data Analytics has access to the transcription files produced by the system, as does BetaNYC as account operator.

NYC AI Use Case Register — Zoom Automated Captions

Available to download

Full event transcripts are embedded in YouTube recordings uploaded to the NYC Open Data Week YouTube channel and are publicly accessible and downloadable.

NYC AI Use Case Register — Zoom Automated Captions

Available to vendor

Zoom, as the technology provider, processes all audio through its platform. BetaNYC, as account owner, also has direct access to transcription files and uses them when editing and uploading recordings.

NYC AI Use Case Register — Zoom Automated Captions

Stored

Stored on 3rd Party Cloud

Transcripts and recordings are stored on Zoom's cloud infrastructure and subsequently on YouTube (Google). Both are third-party cloud platforms outside direct NYC government infrastructure.
Duration: Indefinitely (YouTube recordings remain publicly published; Zoom cloud retention not specified in register)

NYC AI Use Case Register — Zoom Automated Captions

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 UseMeeting attendees should be informed that Zoom Automated Captions and transcription are active during Open Data Week events. The register entry does not describe a formal notice mechanism, but the public nature of the events and their YouTube publication implies participants can reasonably expect to be recorded and transcribed.
Right to Algorithmic TransparencyThe register discloses that the underlying speech recognition model is trained on data proprietary to Zoom, which limits full transparency about how the system works. Members of the public may request information about the system via NYC Open Data channels or by contacting the Office of Technology and Innovation.

Risks and safeguards

Reputational harmAutomated speech recognition can produce inaccurate transcriptions — misidentifying words, names, or technical terms — that are then published publicly on YouTube and attributed to speakers. Errors in a permanent public transcript could misrepresent what a speaker said. No explicit mitigation is described in the register entry; the transcripts are used as-is in the uploaded recordings without a described human review or correction step.