Senators Reintroduce “No Fakes” Bill to Crack Down on AI Deepfakes

U.S. lawmakers are once again moving to confront the explosion of AI‑generated deepfakes, reviving federal legislation intended to shield performers, public figures, and private individuals from unauthorized digital replicas. The bipartisan “No Fakes” Act, reintroduced after failing to advance in the last congressional session, would codify robust rights over a person’s voice and likeness at a time when generative AI can convincingly manufacture speeches, performances, and entire digital personas. As Hollywood and the broader entertainment sector struggle to keep up with fast‑moving synthetic media—and to reconcile it with longstanding norms around contracts, residuals, and consent—the bill’s return reflects mounting pressure to establish national rules before deepfake technologies become even more embedded in day‑to‑day production.

Renewed Push for the No Fakes Act: Rewriting the Rules for AI Deepfakes in Entertainment

Lawmakers from both parties are using the No Fakes Act to signal growing alarm over how AI is reshaping creative and publicity rights. The bill targets the sprawling ecosystem of generative tools now capable of duplicating a person’s face, voice, and performance style with near‑photorealistic precision—often without permission, payment, or transparency. Sponsors say the measure would finally create a clear federal “right of publicity,” equipping both entertainers and ordinary citizens with a direct legal path to challenge deepfake videos, AI‑dubbed performances, synthetic commercials, and songs built on cloned vocals.

The debate is emerging against a backdrop of increasingly visible AI controversies. In the last few years, viral tracks featuring AI versions of famous artists, unauthorized “cameos” by deceased performers in advertising mock‑ups, and hyper‑realistic face‑swaps on social platforms have forced studios, labels, and streaming services to confront how fragile identity has become online. According to a 2023 report from Europol and private researchers, deepfake videos were estimated to be doubling every six months, with entertainment and political disinformation among the fastest‑growing categories.

Crucially, the No Fakes Act does not attempt to ban AI itself. Instead, it aims to impose boundaries on how synthetic media can be used in commercial and entertainment settings. Core components of the proposal would introduce civil penalties for those who knowingly create, distribute, or profit from deceptive AI replicas, while carving out explicit protections for news reporting, documentary work, parody, and other speech shielded by the First Amendment. Supporters believe these protections could bring overdue stability to negotiations over residuals and reuse, at a time when a single likeness scan can be recycled indefinitely.

If enacted, the law is expected to drive substantial changes across the content pipeline, from production contracts to platform moderation:

• Studios: Stricter requirements to obtain written consent for digital doubles, AI de‑aging, and the “resurrection” of deceased performers.
• Artists: Stronger bargaining power over AI‑based reuses of their image and voice, including minimum fees and usage limits.
• Tech Companies: Expanded duties to identify, label, and remove non‑consensual deepfakes, plus clearer record‑keeping for takedown requests.

Building a Federal Shield for Names, Images, Likenesses, and Voices

At the heart of the No Fakes Act is a new federal right enabling performers and public figures to control how their name, image, likeness, and voice are used in commercial contexts—particularly when AI is involved. Today, these rights are governed by a patchwork of state publicity laws and contract provisions, leaving wide gaps that deepfake creators and some tech developers have been quick to exploit.

Under the proposed framework, anyone planning to deploy a digital double, voice clone, or AI‑modeled performance would need explicit, written consent from the individual or their authorized representative. That requirement is designed to shut down scenarios where a performer’s likeness is captured on a single project and then repurposed across multiple titles, languages, or promotional assets without further approval.

Unauthorized uses could trigger meaningful repercussions, including statutory damages and, in extreme cases, criminal liability. Examples of prohibited activity might include:

• AI‑generated songs mimicking a singer’s voice without a license
• Face‑swapped film scenes that insert a celebrity into content they never shot
• Digital endorsements or AI‑narrated ads featuring a recognizable voice without consent

To clarify obligations for both rights holders and distributors, backers of the act highlight three central pillars:

• Mandatory consent for training, modeling, and deploying AI replicas.
• Liability for platforms, producers, and advertisers that knowingly profit from unlicensed deepfakes.
• Safe harbor for news, satire, criticism, documentary projects, and other First Amendment–protected works.

Taken together, these provisions are intended to force studios, advertisers, and AI developers to license identity rights just as they license screenplays, compositions, or character designs. Proponents argue that a uniform federal approach would deter the most exploitative uses of deepfakes, while giving legitimate productions a clear compliance roadmap instead of navigating dozens of inconsistent state statutes.

Hollywood’s Split Response: Creative Promise vs. Legal and Labor Risks

The No Fakes Act has exposed sharp fault lines within the entertainment industry over how aggressively to regulate AI‑generated replicas. On one side, major studios and global streaming platforms view synthetic performers as a way to localize content, extend existing franchises, and reduce expensive reshoots. They are pushing for broad licensing flexibility, along with safe harbors for background or “incidental” uses of AI, such as crowd scenes or minor digital touch‑ups that do not obviously trade on a specific performer’s brand.

On the other side, the technology firms behind leading generative AI models worry that far‑reaching consent requirements and potential statutory damages could chill experimentation. These companies argue that transformative works, satire, and training on publicly available media are already covered by copyright doctrines like fair use, and that new layers of liability might entrench larger incumbents while squeezing startups.

Labor organizations and artist advocacy groups, meanwhile, frame the No Fakes Act as a basic survival framework. Following high‑profile strikes and negotiations in 2023 and 2024 centered on AI clauses, guilds insist that consent, compensation, and audit rights around digital replicas must be non‑negotiable. For them, the prospect of a performer being “recast” by their own AI clone across multiple projects—without ongoing approval—represents an existential threat to human creative work.

Behind the scenes, much of the conversation is practical rather than philosophical: who gets paid, and who carries the legal risk, when AI‑generated doubles or revived performers appear on screen? Union negotiators are pressing for mandatory bargaining over any digital replica, including minimum day rates and residual structures tied to AI‑driven uses. Some producers, by contrast, warn that overly rigid rules could push high‑budget shoots to countries with more permissive AI regulations, echoing earlier offshoring debates around visual effects.

Emerging positions can be summarized as follows:

• Studios: Want predictable licensing frameworks and exceptions for archival footage, local language dubbing, and VFX‑heavy projects.
• Tech companies: Seek protections for AI training data sets, model development, and non‑commercial “sandbox” experimentation.
• Unions: Demand prior consent, fair pay, residuals, and audit rights for any synthetic performance or replicated likeness.
• Independent creators: Worry about navigating expensive legal compliance and stricter platform rules without the backing of major studios.

Preparing for a New Wave of AI Regulation: Practical Steps for Artists, Platforms, and Policymakers

With Congress signaling a more assertive stance on generative AI and deepfakes, stakeholders can no longer rely on informal practices or one‑off deal points. Music labels, talent agencies, guilds, streaming services, and creator platforms are already being pushed toward more formalized, enforceable AI policies.

Rights holders can start by auditing their archives to identify assets that are especially vulnerable to misuse—such as isolated vocal takes, motion‑capture sessions, 3D scans, and behind‑the‑scenes footage. Those materials should be covered by updated licenses and consent forms that explicitly address AI training, simulation, and reuse. In parallel, streaming platforms and social networks that monetize user uploads are under rising pressure to introduce content provenance tools, watermarking or cryptographic labeling, and real-time detection systems for manipulated audio and video.

Industry associations are also working on template clauses that set out when and how AI may be deployed in a project—whether for de‑aging, stunt replication, language dubbing, or promotional content. These model provisions are likely to become standard in film, TV, music, gaming, and influencer agreements.

Key near‑term steps include:

• Codify consent in contracts for any AI training, likeness scanning, or simulated performance.
• Deploy detection tools to identify deepfakes and synthetic audio on platforms and in internal workflows.
• Standardize reporting and appeal processes for performers whose image or voice is misused.
• Educate creators about AI risks and rights before they sign distribution, endorsement, or licensing deals.

For policymakers, the renewed No Fakes Act debate offers an opportunity to decide whether digital replica rules should remain a narrow update to right‑of‑publicity law or evolve into a broader federal regime defining identity in the age of AI. Priority issues include resolving conflicts between state statutes, clarifying jurisdiction for online harms that cross borders, and designing enforcement pathways that individuals and small creators can realistically use, not just well‑funded stars.

Possible approaches under discussion in policy circles include pilot enforcement programs with agencies such as the FTC or DOJ, grant funding for academic research into robust watermarking and provenance standards, and consultations with AI firms and rights organizations to build interoperable metadata frameworks for tracking authorized vs. unauthorized replicas.

Future Outlook: Who Controls Identity When Anyone Can Be Cloned?

As Congress wrestles with generative AI’s impact on creative industries, the reintroduction of the No Fakes Act highlights an urgent attempt to redraw the boundaries of consent, creativity, and control in digital spaces. The bill’s fate—and the scope of any final compromise—will heavily influence how far Hollywood and adjacent sectors can go in adopting synthetic performers without stripping real people of authority over their own image and voice.

Backers of the legislation are betting that bipartisan anxiety over deepfakes is strong enough to move beyond hearings and headline‑driven outrage toward enforceable law. Over the coming months, lobbying by studios, tech platforms, unions, and artist groups will test whether that concern can survive intense industry pushback.

However the details ultimately land, one message is already clear for anyone working in entertainment: the era of “wait and see” on AI‑generated deepfakes is ending. Those who proactively update contracts, workflows, and compliance strategies now will help define how human creativity and machine‑generated media coexist on the next regulatory frontier—and how much power artists retain over their own likeness in a world where almost anyone can be digitally replicated with a click.