Can you use a recording as evidence? How to transcribe it for court

Not legal advice. Admissibility rules vary by state and by court. This is a practical overview for organizing evidence; confirm the rules that apply to your matter with counsel.

A client hands you a phone with two hours of recordings — a heated call, a voicemail, a video from the driveway. Somewhere in there is the thing that matters. Three questions decide whether it helps your case: is it usable, is it accurate, and can you prove it wasn't tampered with?

1. Is the recording admissible?

This is the threshold question, and it's mostly about consent. States fall into two camps:

• One-party consent: a recording is lawful if at least one participant (often your client) consented.
• All-party (two-party) consent: every participant must have consented; a secret recording may be both inadmissible and unlawful.

The wrinkle that costs cases: an improperly obtained recording can taint related evidence and credibility, not just itself. Settle admissibility before you build an argument on top of a clip. When in doubt, treat the recording as a lead to corroborate by other means rather than the centerpiece.

2. Transcribe it accurately — and citably

Courts generally treat the recording as the primary evidence and a transcript as an aid to the trier of fact. You need both. A good evidentiary transcript is more than a wall of text:

• Speaker attribution. Who said what. Disputes over a single line turn on whether it can be reliably attributed.
• Timestamps. Every passage tied back to a moment in the source, so opposing counsel — and the judge — can jump to the audio and confirm it.
• Verbatim fidelity. Including the pauses, crosstalk, and "[inaudible]" markers. Cleaning up speech changes its meaning.

Automated transcription has gotten very good, but "good" isn't "certified." Treat the machine transcript as a fast first pass, then verify against the audio for any line you intend to rely on. The workflow that scales is: transcribe everything, search to find the passages that matter, and human-review only those.

The transcript makes the recording searchable and citable. The recording stays the source of truth.

3. Preserve a chain of custody

Even a perfect transcript fails if opposing counsel can argue the file was altered. Chain of custody is the documented, tamper-evident answer to "how do we know this is the original?" In practice:

• Hash on intake. Compute a cryptographic hash (e.g., SHA-256) the moment the file is collected. If the file changes by a single byte, the hash changes — so a matching hash later is proof it didn't.
• Log every access. Who opened, exported, or annotated the file, and when — in a record that can't be quietly edited after the fact.
• Keep the original untouched. Work from copies; never edit the source. Derived files (clips, transcripts) should point back to it.

How Scribe handles this

This is precisely what Scribe Verbatim is built for. On upload, every file is hashed and entered into a tamper-evident, hash-chained audit log. Recordings are transcribed with speaker attribution and per-line timestamps, and the whole case — recordings, messages, photos, and financial records — becomes searchable in one place, so you find the line that matters in seconds instead of scrubbing through hours. When it's time to produce, you export the originals with a SHA-256 index that proves nothing changed.

You still bring the judgment. Scribe makes the evidence organized, searchable, and verifiable — so the argument is about what was said, not whether you can trust the file.