How to redact audio in body-worn camera footage: names, addresses and radio chatter
Body-worn cameras capture far more than images. During a single police interaction, the microphone may record names, home addresses, telephone numbers, medical information, conversations with victims and witnesses, dispatch communications, officer-to-officer discussions, and radio traffic containing sensitive operational details.
That audio can be essential evidence. It can establish what was said, provide context for an officer's actions, and help prosecutors, defense teams, investigators, and courts understand an incident. At the same time, releasing or sharing an unredacted recording can expose information that has nothing to do with the legitimate purpose for which the footage is being disclosed.
The U.S. Department of Justice's Bureau of Justice Assistance recognizes privacy as an important consideration throughout the BWC lifecycle, including decisions about when footage is viewed, discarded, redacted, and shared. State, local, and tribal laws also vary considerably when agencies respond to public or media requests for body-worn camera footage.
Why audio redaction deserves separate attention
Visual redaction is usually the first concern when an agency prepares bodycam footage for disclosure. Faces, license plates, screens, documents, and other identifying information can be visible throughout a recording.
Audio presents a different challenge because sensitive information can appear for only a few seconds and may not be obvious from the video itself. An officer could ask for a victim's address, a witness could provide a telephone number, or a dispatcher could transmit information about another individual over the radio.
Someone reviewing only the visual track could easily miss these disclosures.
The BJA's body-worn camera guidance emphasizes that agencies need policies covering the access, review, retention, and release of BWC recordings, while model procedures recognize that privacy-sensitive material may need to be redacted before footage is released.
What information might need to be redacted?
There is no universal list that applies identically to every agency or every release. State public-records laws, court orders, agency policy, the circumstances of the investigation, and the intended recipient can all affect what should be withheld.
However, several categories deserve particular attention during an audio review.
Names And Identifying Details
Names may seem harmless when spoken during an ordinary conversation, but context can change their sensitivity.
A victim's name, a juvenile's identity, a confidential source's name, or the identity of an undercover officer can create very different privacy and safety considerations. A person's name may also be accompanied by information such as their address, workplace, date of birth, or relationship to an investigation.
Federal cases involving law-enforcement records have recognized substantial privacy interests in certain personally identifying information, including names and identifying details belonging to agents and third parties.
For that reason, audio review should consider not just individual words but the information revealed by combinations of words.
Home Addresses And Locations
Addresses are another common source of sensitive information.
An officer may ask where someone lives, a witness may give directions, or a dispatcher may repeat an address associated with an incident. In some situations, that information could identify a victim, witness, family member, or other person who is not the subject of the investigation.
The same principle applies to highly specific location details. A recording might reveal a room number, apartment number, workplace, school, shelter, or other location that should not be unnecessarily disclosed.
Telephone Numbers And Other Contact Information
Bodycam conversations can contain telephone numbers, email addresses, social-media usernames, and other contact details.
These may be spoken naturally during an interaction and therefore be much harder to identify than information displayed on a document. A reviewer needs to listen carefully enough to recognize these disclosures and determine whether they belong in the released version.
Medical And Personal Information
Body-worn cameras are sometimes activated during emergencies, arrests, welfare checks, traffic stops, or calls involving people experiencing medical difficulties.
Audio may therefore capture information about injuries, medications, diagnoses, mental-health circumstances, or other highly personal matters. Whether and how such information should be redacted depends on the applicable law and circumstances, but agencies should not assume that information is suitable for unrestricted release simply because it was recorded during an official interaction.
Radio Chatter
Radio communications can introduce an entirely different category of risk.
An officer's bodycam may capture dispatch traffic in the background, including names, addresses, descriptions of suspects, vehicle information, tactical details, officer locations, or information about unrelated calls.
Radio traffic also makes it possible for sensitive information to appear even when the officer wearing the camera is not discussing the subject directly.
This means an audio review should not be limited to conversations between the officer and the people in front of the camera.
Radio chatter can change the context of a recording
One of the most difficult aspects of bodycam audio is that several conversations can happen simultaneously.
An officer might be interviewing a witness while dispatch communications are audible in the background. Another officer may provide information over a radio while a victim is speaking nearby. Multiple voices can overlap, making it difficult to determine which words belong to which conversation.
Simply searching for a person's name is therefore unlikely to be enough for every recording.
Reviewers need to understand the surrounding context and determine whether a piece of audio relates to the incident, another investigation, a protected individual, or ordinary operational communication.
The BJA has highlighted the importance of comprehensive BWC policies covering when recordings can be viewed and shared, while DOJ guidance on video redaction recommends planning for redaction before requests for footage arrive.
Transcription can make audio review more efficient
Listening to hours of bodycam footage from beginning to end is an inefficient way to find every sensitive word or phrase.
Transcription can provide a searchable representation of the audio. A reviewer can search for names, addresses, specific phrases, or other terms and then return to the corresponding point in the recording.
That does not mean transcription should replace listening to the recording. Speech recognition can misinterpret names, accents, overlapping voices, background noise, radio transmissions, and other difficult audio. A transcript should therefore help direct the review rather than become the sole basis for a redaction decision.
A strong workflow combines automated identification with human verification.
Automated redaction needs human oversight
Automation can reduce the repetitive work involved in preparing lengthy recordings, but no system should be treated as infallible.
The U.S. Department of Justice's 2026 Chief FOIA Officer Report provides a useful example: the U.S. Marshals Service reported using automated redaction functionality for body-worn camera footage but noted that the technology was not completely reliable and that human redaction and review remained necessary.
That principle is important for audio because errors can work in both directions.
A system might fail to identify a name that is spoken unclearly. It might also flag an innocent occurrence of a word that happens to match a search term. A final reviewer needs to determine whether the proposed redaction actually protects sensitive information without removing evidence unnecessarily.
Preserve the original recording
Redaction should take place on a controlled copy rather than destroying or overwriting the original evidence.
The source recording may be needed later for prosecution, defense review, litigation, an internal investigation, or another authorized purpose. Agencies therefore need to maintain appropriate controls over the original while producing a separate version for disclosure.
The DOJ's BWC policy guidance specifically addresses safeguarding downloaded recordings against tampering or deletion, while federal agency procedures demonstrate the importance of maintaining controlled copies when particularly sensitive law-enforcement information is involved.
A sensible workflow is therefore:
Original recording → controlled working copy → audio and video analysis → redaction → quality review → approved disclosure copy
The exact technical implementation will depend on the agency's evidence-management environment, but the underlying principle remains the same: privacy editing should not compromise evidence preservation.
Build audio redaction into the disclosure process
Redaction should not be treated as something that happens only after someone asks for a recording.
The DOJ's best-practice guidance recommends considering video redaction requirements when agencies first implement recording technology and identifying the necessary tools, people, and resources before disclosure requests arrive.
That approach makes sense for audio as well.
Agencies can establish categories of information that require careful review, define who can approve a redacted recording, document the reason for significant redactions, and establish procedures for securely delivering the final copy.
This also helps distinguish between different uses of the same footage. A recording prepared for public release may require substantially more privacy protection than a version shared internally with authorized investigators.
Combining audio and video redaction
Bodycam footage should ultimately be treated as a combined evidence source rather than two completely separate files.
A face could remain visible while the person's name is spoken in the audio. Conversely, an individual's identity might be hidden visually while a radio transmission reveals their address.
That creates a straightforward risk: an apparently well-redacted video can still disclose someone's identity through another part of the recording.
Our Secure Redact technology addresses this by bringing audio and video redaction into the same workflow. Automated transcription can help identify sensitive spoken content, while visual detection and tracking can address faces and other identifiable information, allowing agencies to review both channels instead of treating audio as an afterthought.
Discover more about Secure Redact, and find out about starting a free trial.
A practical audio redaction workflow
For agencies processing bodycam footage regularly, a repeatable process can make reviews more consistent.
1. Identify the purpose of disclosure - Determine whether the recording is being prepared for a public-records request, litigation, prosecution, training, an internal investigation, or another authorized purpose.
2. Preserve the source - Secure the original recording and establish a controlled copy for processing.
3. Identify sensitive categories - Consider names, addresses, telephone numbers, medical details, juvenile information, confidential sources, investigative details, and radio traffic.
4. Transcribe or search the audio - Use available technology to locate potentially sensitive words and sections, while recognizing that automated transcription can make mistakes.
5. Review the surrounding context - Listen to the relevant section rather than relying solely on a keyword hit. Determine who is speaking, what the information refers to, and whether it actually requires redaction.
6. Apply the redaction - Remove or obscure the appropriate section of audio while preserving enough surrounding material to maintain the evidentiary context.
7. Review the complete disclosure copy - Check that sensitive information has not been missed and that the redaction does not unnecessarily remove relevant evidence.
8. Document and securely release - Record the appropriate processing and approval information, then provide the final copy through an authorized method.
Redaction should protect privacy without destroying context
The purpose of audio redaction is not to make a recording silent whenever sensitive information appears. Excessive editing can make conversations difficult to follow and potentially remove context that is important to understanding an incident.
The objective is targeted protection.
If an address is spoken during an otherwise relevant conversation, the address may need to be protected while surrounding speech remains available. If a radio transmission contains sensitive information about an unrelated investigation, only the relevant portion may require attention.
That balance allows agencies to protect victims, witnesses, officers, suspects, and other members of the public while preserving the evidentiary value of the recording.
Better audio review starts before disclosure
Body-worn camera programs create enormous quantities of information, and audio adds a layer of sensitivity that can easily be overlooked when teams concentrate primarily on faces and other visual identifiers.
Names, addresses, medical details, contact information, confidential sources, and radio communications can all appear in seconds. A reliable process therefore combines clear agency policy, controlled evidence handling, automated assistance where appropriate, and human quality assurance.
The BJA's guidance emphasizes that privacy considerations should be incorporated into the wider BWC program rather than addressed only at the point of release.
For agencies processing substantial volumes of footage, the practical goal is straightforward: find sensitive information efficiently, redact only what needs protection, preserve the original evidence, and make sure the final recording is genuinely ready for its intended recipient.
Frequently asked questions
-
Yes. Audio and video can be processed separately, although agencies should review both channels together before releasing the final recording.
-
Depending on the circumstances, searches may include names, addresses, telephone numbers, medical terms, identifying details, confidential-source information, and other terms specified by agency policy.
-
No. Radio traffic should be assessed based on its content and the applicable disclosure requirements. Some communications may be relevant to the incident, while others could contain sensitive operational or personal information.
-
No. Transcription systems can make errors, particularly with unusual names, poor audio quality, overlapping speech, accents, or background noise. Human review remains important.
-
The original should generally be preserved separately from any redacted disclosure copy. Agencies should follow their evidence-management policies and applicable legal requirements.
-
That depends on the agency and the purpose of disclosure. Agencies should establish clear authorization procedures identifying who can review and approve recordings before release.
-
It can if too much material is removed or if important context is obscured. Targeted redaction and careful quality review can help protect sensitive information while retaining relevant evidence.