Litigation and legislation around AI and copyright have been slowly heating up over the last few years, but recently, the U.S. Senate read a bill sponsored by presidential allies with bipartisan support.
As copyright enforcement shifts out of legal teams and into the systems that manage content, rights, and distribution, DAM is becoming the control point. Echoing conservative copyright law in the EU, DAMs will have no choice but to enforce copyright licensing — and they should start now.
Many great minds have talked about copyright and AI (Lessig, Samuelson, Butterick, and Gervais). I’m not one of them, but I study. Let’s start with an easy definition: copyright is about the right to make copies. Remembering that the definition is in the name helps me when I administrate copyright issues in a DAM system, where those rights are tracked and enforced.
Or rather, let me be honest. I’ve worked in digital content marketing since Yahoo directories, and I’ve rarely heard stakeholders express concern about copyright infringement. However, I know of cases in which failing to protect an individual caused large settlements and stress — and those risks are becoming harder to ignore as AI accelerates content creation and reuse.
Your customers search everywhere. Make sure your brand shows up.
The SEO toolkit you know, plus the AI visibility data you need.
Start Free Trial
Get started with
Why AI copyright risk is escalating now
As the safekeepers of metadata, licensing, and brand safety, DAM managers and administrators are also gatekeepers of licensed intellectual property. The worry now is about who owns and controls AI and LLMs.
About 40 copyright claims have been filed against generative AI (genAI) and LLM providers over the past three to six years, including OpenAI, Anthropic, Stability AI, and Midjourney. Many high-profile class actions were consolidated in the spring of 2025 under In re OpenAI, Inc. Copyright Infringement Litigation.
The European Parliament also discussed a rising number of claims in its recent resolution, released in March 2026, “Copyright and generative artificial intelligence – Opportunities and Challenges (P10_TA (2026)0066),” noting the “widespread evidence of violations of copyright rules by genAI providers.”
Naughty genAI isn’t the only thing the EU Parliament took note of. They also said years of abuse by ISPs, tech providers, data brokers, and others led to their resolution that every piece of data belonging to an individual — including their likeness and voice — is intellectual property. However, they’ve stopped just short of making genAI technology illegal, even though it poses a high risk of copyright infringement.
Instead, genAI providers, ISPs, and software, including DAM platforms, are instructed to follow the EU’s Code of Best Practice Guidelines. Best practice is to have rights documentation signed by the subject, with legal representation present. Documents should include a broad representation of the types of assets in which an individual’s voice and/or likeness could be used. If the caseload concerning damages against genAI doesn’t lighten, Parliament will make the resolution law.
No fakes, no frauds: Why U.S. DAMs must enforce copyright
U.S. legislators have recently introduced two bills concerning the use of genAI. The No AI Fraud Act (H.R. 6943) would provide protection for individual voice and likeness in assets created with genAI. Previous law made individual likeness and voice the IP of only celebrities and public figures. This would extend rights to every individual. Even though our marketers and others are using DAM assets for commercial purposes, that use isn’t outside the scope of the act. The bill failed to reach the Senate but was incorporated into another measure.
The No Fakes Act (S.4875) provides for the protection of individual voice and visual likeness in genAI, especially from what are known as AI fakes or deep fakes. Deep fakes are created by deep learning (DL) engines to create highly realistic, forged content. The bill’s full title is the “Nurture Originals, Foster Art, and Keep Entertainment Safe Act.” It provides for the terms of an individual’s copyright (life plus 70 years), and the term duration of transferred licenses. Each license must be signed in writing and include a description of the intended use of the digital copies.
Damages for publishing an unauthorized digital replica of an individual’s IP could cost $50,000 — per copy — or the cost of actual damages determined by a court, and $5,000 per copy of an individual’s voice.
What should DAMs do?
A few practical steps can reduce the risk.
- Document individual rights: Ensure everyone photographed or recorded has signed a transfer-of-rights agreement covering all assets in which they appear. The EU has adopted an opt-out model for marketing. Consider adopting such a model for photoshoots and recording sessions.
- Govern AI LLMs: Create metadata to tag assets that contain individuals’ likenesses and voices. Then create a workflow that automatically recognizes faces, bodies, and voices to exclude them from your AI training. Next, make sure your DAM can spot AI fakes and decide what to do with them. Your organization might want to create a policy forbidding the creation of AI fakes, since there may be liability issues.
- Audit for assets lacking copyright protection: AI-generated content can’t be copyrighted. Competitors and the public can legally take it and reuse or remix it, and your company will have no legal recourse. If you can’t track down licenses, delete them, or at least archive them.
- Tag all AI assets: The actual tag name may differ depending upon your use of standards, but three main categories are important: AI-generated, or genAI, AI-assisted, and human-created.
- Train production on human-created genAI processes: These assets, generated by prompts and using your brand’s asset, copy, or logo, may be the only ones that are protected IP. Consult with your legal department before drafting any AI governance, and ensure you have sign-off on your training.
Copyright governance is now an ongoing practice
There will never be anything easy about copyright and intellectual property, but metadata, workflows, and governance will go a long way. Make your copyright governance a topic for bimonthly or quarterly review, and consider creating a community of practice within your organization to serve as your advisory team.
