Your Catalog Is an AI Asset. Do You Know What You Own?
Your Catalog Is an AI Asset. Do You Know What You Own?
The news out of Hollywood over the past year has a consistent throughline: legacy content is being repriced in real time.
Lionsgate gave Runway access to its 20,000-title library to train a proprietary video generation model. Disney structured a $1 billion arrangement with OpenAI covering over 200 characters and a significant portion of its IP portfolio. Universal, Sony, and Warner have each signed licensing deals with generative music platforms. Across every media category, the same thing is happening: studios and labels are monetizing their catalogs in a way that simply wasn't possible three years ago.
If you're an independent producer or a smaller rights holder, the temptation is to watch these deals from the sidelines and assume they don't apply to you. That instinct is wrong.
What the Big Deals Actually Establish
The Lionsgate-Runway structure is instructive. The deal is proprietary — Runway can't use the Lionsgate catalog to build a public model. But that restriction is also a feature. Lionsgate gets a custom AI tool trained on its own aesthetic, built at someone else's cost, while retaining exclusivity over the output. The catalog is the asset. The AI company is, in effect, paying for access to it.
The Disney-OpenAI arrangement tells a slightly different story. At $1 billion, it's as much a strategic investment as a licensing deal, but the underlying logic is the same: decades of IP have quantifiable value to companies that need high-quality, rights-cleared content to train on.
What these deals collectively establish is market structure. The terms being negotiated now — what constitutes "training use," how revenue from downstream AI outputs gets shared, what restrictions govern model deployment — will become standard reference points. Early movers are writing the playbook. Everyone who comes after negotiates against it.
What This Means for Indie Rights Holders
Independent producers often think of their back catalogs as passive assets: something that generates a little VOD revenue, maybe gets licensed for a remake or a format sale. AI changes the calculus in a few ways.
First, demand for rights-cleared content for AI training is real and growing. AI companies need content they can legally license, and a library with clear chain of title and well-documented rights is genuinely valuable — not just to studios building internal tools, but to AI platforms that want to offer licensed training datasets as a service. Catalog aggregators are already forming for this exact purpose, operating similarly to performance rights organizations but focused on AI training territories.
Second, the valuation lens is shifting. Music catalog multiples — already running 10 to 18 times net publisher share in recent years — now have an AI licensing upside component baked into buyer modeling. Film and TV catalogs are being evaluated similarly. If you're sitting on a library and thinking about financing, a sale, or a partnership, the people on the other side of that table are running AI scenario models on your content.
Third, if you haven't already, you need to look hard at what rights you've already licensed away. A lot of standard distribution agreements, particularly deals done before 2022, contain broad language around "all media, formats, and technologies now known or hereafter devised." Courts and legal teams are still working out whether that language covers AI training use. In some cases, it clearly does. In others, it's genuinely ambiguous. But you don't want to find out during a due diligence process or after a deal has closed.
The Rights Question Nobody Is Asking Early Enough
Here's the practical issue: most independent producers haven't done an AI-specific audit of their library deals. They know what distribution rights they've granted, roughly. They may know where their chain of title is clean or fragile. But they haven't mapped their existing agreements against the specific question of AI training use.
That audit is not complicated, but it needs to happen before someone else does it for you. Specifically:
Look at your distribution agreements, particularly any with broad technology grants. Look at your underlying acquisition agreements — what did you represent and warrant to your distributor about scope of rights? Look at any co-production or financing agreements that include IP provisions.
If you have music in your catalog, separate out the sync licenses from the master recording rights, and check what language covers downstream uses. Music is the most litigated area right now, and the deals being done by the majors don't automatically flow downstream to independent rights holders who've licensed to them.
If you're an indie producer with a library, the moment to get ahead of this is before you're in active discussions with an AI platform, a studio, or a buyer who's doing their own due diligence. The rights questions are solvable. They're just a lot easier to solve when you're not in the middle of a transaction.
The catalog you've been treating as a slow-burning asset may be worth more than you think. The question is whether you can actually access that value when the time comes.