When Danish outlet Zetland spun off Good Tape as a standalone transcription service in 2023, it made a bold claim: this would be transcription software that journalists could actually trust with their sources. That’s no small promise in an era where tech companies routinely harvest user data for AI training, potentially exposing confidential sources or handing competitive advantages to rivals. The question isn’t whether Good Tape delivers accurate transcription—it does—but whether newsrooms can trust it with their most sensitive material.
What do 1,000 journalists and PR pros know about AI that you don't? They took AI Quick Start, a 1-hour live class from The Media Copilot. 94% satisfaction. Find out how to work smarter with AI in just 60 minutes. Next class May 8. Get 20% off with the code AIPRO: https://mediacopilot.ai/
Key Takeaways
- Good Tape, spun out of Zetland, markets itself as journalist-trustworthy.
- Company commits to not training on user audio, key for source protection.
- Still verify encryption, retention and processor terms before standardizing.
The stakes are particularly high for journalists. A leaked recording can destroy careers, endanger sources, and compromise months of investigative work. Meanwhile, transcription services that train their AI models on user content essentially turn journalists into unpaid data providers, feeding proprietary interviews into systems that competitors might later access. Jacob Granger, senior reporter at journalism.co.uk, put it bluntly: “If we’re feeding our transcripts into untold other generative AIs and they use that to train their model, we’re giving away a very valuable part of our work to benefit them.”
Good Tape’s approach to these concerns reflects its newsroom origins. The company hosts its infrastructure on EU-based servers, subjecting itself to GDPR and other European privacy regulations that impose stricter controls than U.S. data protection laws. But server location is just the beginning of the security equation. What matters more is how the company handles the data once it arrives, who has access to it, and what happens to it after transcription is complete.
Risks identified in Good Tape’s security posture
The primary risk with any transcription service centers on data exposure—whether through breaches, employee access, or AI training practices. Good Tape addresses the AI training concern directly: the company states it never uses customer transcripts to train its models. This differs markedly from competitors like Otter.ai, which has acknowledged using de-identified user recordings for model improvement, raising questions about whether true de-identification is even possible with voice data.
Another risk involves data retention and deletion. While Good Tape allows users to delete their transcripts from the platform, the documentation doesn’t specify retention periods for audio files or whether deletion is immediate and permanent across all backup systems. For journalists working on time-sensitive investigations or with whistleblowers, understanding exactly when and how completely their data disappears matters. The platform does offer an option to process audio without saving it—users can uncheck a box during upload to prevent audio storage—but this security-conscious feature isn’t prominently advertised.
Security controls Good Tape has implemented
Good Tape’s security framework centers on AES-256 encryption, the same standard used by the U.S. government for classified information. This encryption applies to data both in transit and at rest, meaning files are protected during upload and while stored on servers. The EU server location adds a regulatory layer of protection—European privacy laws require explicit consent for data processing and impose substantial fines for violations, creating financial incentives for compliance that don’t exist in less regulated jurisdictions.
The company’s most significant security feature may be its business model. Unlike free or ad-supported transcription services that must monetize user data somehow, Good Tape operates on straightforward subscription pricing: $17 monthly or $190 annually for 20 hours of transcription. This removes the financial pressure to extract value from user content through AI training or data brokering. CEO Tav Klitgaard emphasized this philosophy when discussing confidential sources: “It cannot leak and you cannot train any models on it.”
The platform’s authentication and access controls remain less documented. While the service requires user accounts and passwords, the available documentation doesn’t specify whether it supports two-factor authentication, single sign-on for enterprise customers, or role-based access controls for newsroom teams. These features become critical when multiple journalists share an organizational account or when newsrooms need to comply with their own security policies. The platform does maintain file organization through a sidebar system that could theoretically support user permissions, but current documentation doesn’t confirm this capability.
Security checklist for Good Tape users
Before trusting Good Tape with your newsroom’s sensitive transcripts, verify the following:
- Does your organization require SOC 2 Type II compliance?
- Do you handle data subject to GDPR/CCPA?
- Do you need data residency in specific geographic regions?
- Are you subject to industry-specific regulations (HIPAA, FERPA, etc.)?
- Do you require custom data processing agreements?
- Do you need detailed audit logs of all data access?
- Does your IT department require SSO integration?
For journalists handling particularly sensitive material, Good Tape’s option to process audio without storage provides an extra security layer—though users must remember to actively select this option with each upload.
Newsrooms should evaluate Good Tape against their specific threat models and compliance requirements. For many journalists, the combination of encryption, EU hosting, and no AI training on user data will meet their security needs. Others may require additional documentation about audit logs, incident response procedures, or enterprise security features before committing sensitive interviews to any third-party platform.
Frequently Asked Questions
Good Tape is considered one of the more trustworthy transcription options for journalists due to its automatic audio deletion policy and GDPR compliance under Danish law. For interviews involving sources at risk of physical harm or legal jeopardy, journalists should consult their newsroom’s security team before uploading recordings to any cloud service.
Good Tape automatically deletes your original audio files from its servers after the transcription process completes. This is a deliberate privacy feature—Good Tape does not retain your recordings after processing, significantly reducing risk if the service were ever subject to a data breach or government request.
Yes. Good Tape is GDPR-compliant and operated under Danish jurisdiction with strong EU data protection regulations. The company provides data processing agreements (DPAs) that newsrooms can sign to formalize compliance requirements—important for EU-based news organizations with formal data protection obligations.
Yes. Good Tape provides documentation on its data processing and security practices. Newsrooms can request DPAs and technical security questionnaire responses. Reputable news organizations typically require vendors to complete security assessments before approving any tool for sensitive editorial workflows.
For the highest-security transcription needs, local offline tools eliminate cloud risk entirely. Options include running OpenAI’s Whisper model locally on an air-gapped computer or using locally installed transcription software. For most journalistic purposes Good Tape’s privacy protections are adequate, but truly sensitive national-security-level interviews may warrant offline processing only.
