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 Meeting Transcription for Vendor Trainings

What it is for

Inform
  • Transcripts of vendor training sessions are shared with all registered participants after the meeting and made available to anyone who requests a link to the recording.
Run by
Mayor's Office of Contract Services (MOCS)
Where
No fixed location
Kept
Not stated by the NYC Office of Technology and Innovation.
Shared with
Accountable organization, Download, Vendor

What it is for

Inform

This system uses an AI feature built into the Cisco Webex platform to automatically transcribe spoken words during public vendor training sessions hosted by the Mayor's Office of Contract Services. After each meeting, the transcript is shared with attendees and made available to anyone who requests access to the recording. The transcription is fully automated with no human editorial review before distribution. Participants should be aware that everything spoken during these sessions is captured and converted to text.

What it collects and what happens to it

Data taken in

Biometric
Identifiable data
  • The spoken words of all meeting participants are captured as audio input. The register states that the words spoken during a meeting are captured by the tool, which means identifiable voices of named participants are processed.

Processing

Speech & Audio
  • The AI transcription feature uses speech-to-text processing to convert spoken audio from Webex meetings into a written transcript. This is the core processing technique described in the register entry.

What it does

Sensing (Perceptive AI)
Autonomous
  • The system autonomously listens to spoken audio during live Webex meetings and converts speech to a structured text transcript without human intervention during the conversion process.

Outputs

Generated content
Identifiable data
  • The tool produces a text transcript of the words spoken during the meeting. Because the transcript captures the speech of identified participants in a named session, the output is identifiable in context.

Run by

Mayor's Office of Contract Services (MOCS)
  • The Mayor's Office of Contract Services (MOCS), specifically its Learning & Development department, deploys this AI transcription tool during biweekly vendor training sessions hosted on Webex.

NYC Algorithmic Tools Compliance Report

Built by

Cisco
  • Cisco provides the AI transcription feature as a built-in capability of the Webex platform. The tool is off-the-shelf and the underlying training data is proprietary to Cisco.

NYC Algorithmic Tools Compliance Report

Kept for

Not stated by the NYC Office of Technology and Innovation.

Shared with

Available to the accountable organization
  • MOCS retains access to the transcript and recording link, which it shares with attendees and upon request.
Available to download
  • A public link to the recording and transcript is shared with all who register for or attend the training, and is available to anyone who requests it.
Available to vendor
  • As an off-the-shelf Webex feature, Cisco processes the audio and may have access to transcription data as part of its platform operations. The training data for the underlying model is proprietary to Cisco.

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 UseParticipants in MOCS Webex training sessions should be informed that AI transcription is active during the meeting. The transcript is shared publicly with all registered attendees and is available upon request, so attendees are notified post-meeting at minimum. Contact the Mayor's Office of Contract Services for further information.
  • Right to Algorithmic TransparencyThe system uses a proprietary AI model built into Cisco Webex. The specific algorithm and training data are not publicly disclosed by Cisco. The NYC Algorithmic Tools Compliance Report provides the agency's description of how the tool is used. Members of the public may consult the register at https://data.cityofnewyork.us/City-Government/Algorithmic-Tools-Compliance-Report/jaw4-yuem for details.

Risks and safeguards

  • Psychological harmParticipants may feel surveilled or inhibited knowing their spoken words are automatically transcribed and made publicly available.Safeguard: MOCS shares transcripts only of public vendor training sessions (not sensitive internal meetings), and the public nature of the sessions is established in advance through registration. Participants should be clearly notified of transcription at the start of each session.
  • Societal & cultural harmSpeech-to-text AI systems are known to perform less accurately on non-native accents, certain dialects, and non-English speech, potentially producing inaccurate transcripts that misrepresent what participants said.Safeguard: The register does not describe any accuracy validation or correction mechanism. MOCS should establish a process for participants to flag and correct errors in distributed transcripts.