← All tool ratings

Descript

AI-powered audio and video editing with text-based editing.

Adequate
https://www.descript.com Reviewed 2026-04-02 Editorial assessment by Mike Schneider — not an independent security audit

What should journalists know about Descript?

Descript's text-based editing genuinely changes how you work with audio and video — edit a transcript like a Google Doc and the media follows. The AI features (filler word removal, eye contact correction, Studio Sound noise reduction) save hours on deadline. But your media uploads to Descript's cloud for processing. For a podcast about public policy, fine. For audio of a confidential source describing corporate fraud, no. Descript is independently owned (not Spotify — that was rumor, not fact), VC-backed with $101M raised, and now doing $55M ARR with 75% year-over-year growth under new CEO Laura Burkhauser. The product is excellent for non-sensitive production. Just know what you're uploading and where it goes.

Best for

Podcast editing, video journalism, interview tape cutting, content repurposing, fast-turnaround spoken-word content. Non-sensitive media production where speed matters more than operational security.

Not for

Editing audio or video containing confidential source material, whistleblower recordings, or any content where the existence of the recording itself is sensitive. Use DaVinci Resolve (free, local processing) or Audacity for that.

Security & Privacy

Encryption in transit Yes

Data is scrambled while being sent to their servers

Encryption at rest Yes

Data is scrambled when stored on their servers

Data jurisdiction United States (AWS infrastructure)

Where servers are located — affects which governments can request your data

Security rating Adequate

Privacy policy summary

Descript processes media on cloud servers (AWS). SOC 2 Type II compliant, GDPR and CCPA aligned. States it does not use your content to train production AI models. Internal R&D models only use data from users who opt in — opt-in is disabled by default. Voice cloning training audio may be used to improve their technology after being disassociated from your account. Descript employees may listen to audio samples for internal QA. Has agreements with sub-processors (including OpenAI) that your data won't be used for third-party model training. Content retained until you delete your project.

How to protect yourself:

For non-sensitive content, Descript is fine as-is. For somewhat sensitive material: (1) delete projects immediately after exporting final cuts, (2) never upload raw source audio that includes identifying information about confidential sources, (3) disable the data-sharing opt-in in settings, (4) use DaVinci Resolve (free, fully local) or Audacity for anything sensitive. For voice cloning: understand that your consent recording and training audio may be used for Descript's R&D even after disassociation from your account.

SOC 2 Type II compliant. AES-256 at rest, TLS 1.2 in transit. GDPR/CCPA aligned. AI training opt-in disabled by default. No breaches on record. Rating reflects cloud-only processing model — all media must leave your machine. Fine for non-sensitive production; not appropriate for confidential source material.

Who Owns This

Owner Descript, Inc. (independent, private). Founded 2017 by Andrew Mason (former Groupon CEO). Mason stepped aside to Executive Chairman in 2025; Laura Burkhauser (ex-Twitter Director of Product) became CEO.
Funding VC-backed. $101M raised across 4 rounds. Series C ($50.6M, Nov 2022) led by OpenAI Startup Fund. Other investors: Andreessen Horowitz, Redpoint Ventures, Spark Capital. $55M ARR as of 2025 with 75% YoY growth. No Spotify acquisition — that was industry rumor, never materialized.
Business model Freemium SaaS. Free tier for adoption; paid tiers gate AI credits, media minutes, 4K export, and team features. September 2025 pricing overhaul introduced metered usage (media minutes + AI credits) replacing flat transcription hours.

Known issues

Consumer Reports (March 2025) assessed Descript's voice cloning safeguards as 'somewhat effective' — better than most competitors (requires recording a consent statement) but bypassable via cloning the consent statement through another service. Pricing overhaul in September 2025 forced migration to metered media minutes + AI credits, frustrating some long-time users on legacy plans. Voice cloning training audio is used for R&D after disassociation from accounts. OpenAI is both an investor and a sub-processor — potential conflict of interest worth noting. No public data breaches or security incidents found.

Pricing

Free (60 media minutes/month, 100 one-time AI credits). Hobbyist: $16/month (annual) or $24/month. Creator: $24/month (annual) or $35/month. Business: $50/month (annual) or $65/month. Enterprise: contact sales. Pricing overhauled September 2025 — now meters both media minutes (uploads/recordings) and AI credits (Studio Sound, Eye Contact, filler removal, etc.). Unused minutes and credits do not roll over.

None known.

This is an editorial assessment based on publicly available information as of 2026-04-02, using our published methodology. Independent security review is pending. Security posture can change at any time. This is not a guarantee of safety.

Something wrong or outdated? Report it.