Box Commons is a data cooperative and standards body uniting independent broadcasters, podcasters, and voice artists to ensure ethical, consent-based audio licensing in the AI era.
AI models are moving beyond text. They need to hear real conversations, real accents, real rooms. For AI developers, podcasts, community radio broadcasts, and voice reels are a goldmine of natural speech. But for independent creators, this unregulated extraction is an existential threat to livelihoods and rights.
Audio is one of the fastest-growing segments. The recordings independent creators delete every quarter to save storage are exactly what this market is buying.
The music industry has SoundExchange. Union actors have SAG-AFTRA. Independent podcasters, radio stations, and voice artists have no collective representation, no standard, and no leverage.
The insurance industry has decided: if you can't prove where your audio data came from and who consented, it's too risky to cover. Certification is the price of entry.
We don't just advocate for ethical AI; we provide the infrastructure to guarantee it. When audio carries the BC-Certified seal, AI companies and researchers know it meets the highest standards of ethical sourcing. Think of it as a nutrition label for audio data. Here is how our four pillars protect your work.
You are in control. Our standard ensures that AI models only train on your audio if you have explicitly opted in. No loopholes, no hidden terms of service. Consent is granular, so you can authorize transcription while restricting voice cloning. And it's revocable: change your mind at any time.
We separate your voice from the background. Our protocol allows you to license specific elements of your audio, like a spoken-word track, without compromising the rights to underlying music, sound effects, or syndicated content. Only what you own gets packaged.
Think of this as a nutrition label for your audio. We embed clear, standardized information directly into your files: who recorded it, where, when, and exactly what permissions were given. Anyone who accesses the file knows precisely what they can and cannot do.
Using secure, industry-standard digital fingerprints, we create a permanent record for your audio. If someone tries to strip away your ownership details, the file breaks. Your rights travel wherever your audio goes.
Box Commons is not a tech platform. It is a 501(c)(6) member-owned cooperative. Our strength comes from uniting the diverse voices that make up the independent audio landscape under a shared structure where every participant has an equal say.
Whether you run a community radio station or a local faith-based network, your archives contain decades of valuable, natural speech. We protect your legacy broadcasts from unauthorized scraping and provide a pathway to safely license your archives to sustain your station's future.
From seasoned voice actors to college students recording their first interview, your voice is your livelihood. We give you the tools to retain total control over your vocal likeness and participate fairly in the AI economy, with collective leverage no individual creator can build alone.
We bridge the gap between ethical engineering and creator rights. By partnering with Box Commons, you gain access to high-quality, legally compliant audio data, allowing you to build the next generation of AI models without the risk of copyright infringement or ethical compromise.
Our position paper lays out the audio data gap, the shift from scraping to licensing, the consent problem, the cooperative model, and the insurance imperative. It is the most comprehensive look at what we are building and why.
Clean Audio Data: A Framework for Ethical Sourcing, Consent, and Provenance in AI Training
Read the position paper →Tech platforms have organized to extract the value of our audio. It is time we organize to protect it. Box Commons is a member-driven data cooperative fighting for ethical AI standards and creator compensation. Add your name to the movement.