The unauthorized use of Scarlett Johansson’s likeness in AI-generated advertisements brought the issue of non-consensual deepfakes into the global legal spotlight. Effectively addressing these violations requires an understanding of existing personality rights, the evolving landscape of digital privacy, and the specific mechanisms individuals can employ to mitigate the unauthorized use of their biometric or digital identity.
The Intersection of AI Technology and Personality Rights
Deepfake technology utilizes sophisticated machine learning algorithms—specifically Generative Adversarial Networks (GANs)—to map the facial expressions, voice, and mannerisms of a target individual onto another body or to generate entirely synthetic, realistic personas. When these tools are applied to high-profile figures like Scarlett Johansson without consent, they do not merely represent a privacy breach; they constitute a commercialization of identity.
The primary legal tension in these cases involves the collision between free expression and the "right of publicity." In many jurisdictions, individuals possess an inherent right to control the commercial use of their name, image, and likeness (NIL). When an AI application or advertisement utilizes these elements to imply endorsement or to misappropriate someone's persona for profit, it violates these established civil rights.
Legal Frameworks and Challenges
Current legal systems are often playing catch-up with the speed of AI advancement. However, several pillars of law provide pathways for recourse against non-consensual deepfakes:
- Right of Publicity: This is the most significant legal hurdle for AI platforms. It grants individuals the right to prevent others from using their identity for commercial purposes without permission. While laws vary significantly by state and country, the unauthorized use of a celebrity’s image to sell a product is widely recognized as an actionable tort.
- Defamation and False Light: If an AI-generated video misrepresents an individual in a way that is harmful to their reputation or casts them in a light that is highly offensive to a reasonable person, the victim may pursue claims of defamation or false light invasion of privacy.
- Copyright Law: While an individual’s face itself is not copyrighted, the underlying photography or videography used to "train" the AI model often is. Plaintiffs may argue that the unauthorized use of their proprietary film footage or photos to train a generative model constitutes copyright infringement.
- Legislative Updates: Since 2023, several legislatures have begun drafting or enacting specific statutes targeting digital replicas. These laws seek to clarify that "digital clones" fall under the same protections as physical likenesses, effectively closing loopholes that AI developers previously exploited.
Verification: Identifying AI-Generated Content
For the average consumer and for legal teams monitoring potential misuse, identifying deepfakes has become a necessary technical skill. While AI detection models are improving, they are rarely 100% accurate.
- Artifact Detection: Look for inconsistencies in the "seams" of the video. This includes blurred edges around the hair, mismatched skin tones between the neck and face, and unnatural blinking patterns.
- Audio-Visual Desynchronization: AI-generated speech sometimes fails to perfectly align with the subtle movements of the tongue and lips. If the audio seems "disconnected" from the physical performance, it is a primary indicator of a synthetic overlay.
- Digital Forensic Tools: Organizations and individuals are increasingly using professional forensic tools to analyze pixel-level anomalies or discrepancies in metadata that distinguish synthetic content from raw camera footage.
Table: Comparison of Legal Remedies for Digital Impersonation
| Legal Path | Scope of Protection | Primary Requirement for Success |
|---|---|---|
| Right of Publicity | Protects commercial use of image/voice | Proving commercial misappropriation |
| Defamation | Protects reputation and public image | Proving the content is false and harmful |
| Copyright Infringement | Protects specific creative assets | Proving ownership of source material |
| Statutory Digital Replica Laws | Protects "digital doubles" | Proving non-consensual simulation |
Practical Steps for Protecting Digital Identity
If you are an individual facing the unauthorized use of your likeness, or if you are interested in the strategies public figures utilize, the following steps are considered industry standard in the current 2026 landscape:
- Cease and Desist: The first step is almost always a formal legal demand sent to the host platform (e.g., social media sites) and the creator of the deepfake. Most major platforms have "Intellectual Property" or "Impersonation" reporting portals that allow for the expedited removal of content that violates their terms of service.
- DMCA Takedowns: If the AI content uses copyrighted images or clips to train the model or to display the fake, a Digital Millennium Copyright Act (DMCA) notice is an effective tool to compel service providers to remove the content.
- Digital Watermarking and Registry: Many public figures are now adopting "proof of humanity" tools, such as blockchain-based registries or invisible digital watermarks, which certify their official content. This makes it easier for platforms to distinguish between authorized media and unauthorized clones.
- Strategic Litigation: In extreme cases, such as those involving significant financial loss or severe reputational damage, filing a lawsuit for damages acts as a deterrent. These cases often set legal precedents that clarify how courts interpret "fair use" regarding generative AI.
The Role of Platforms and AI Developers
The burden of responsibility is shifting. In 2026, technology companies that provide deepfake tools are increasingly expected to implement "safety rails." These include:
- Mandatory Disclosure: Requiring that any AI-generated video feature a permanent, non-removable watermark indicating that the content is synthetic.
- Biometric Filtering: Implementing database checks to prevent users from uploading or using the likenesses of known public figures, a feature that significantly reduces the occurrence of high-profile deepfakes.
- Cooperation with Law Enforcement: Providing audit logs when a specific tool is used to generate malicious or illegal content, ensuring that the developers of these tools can be held accountable if they facilitate systematic abuse.
Common Misconceptions Regarding Deepfakes
A frequent error in public perception is the idea that "everything on the internet is fair game." This is legally inaccurate. The assumption that because a person is famous, their image is in the public domain, is a misunderstanding of both copyright and publicity rights. Even public figures retain the right to control how their persona is utilized in a commercial context.
Another common myth is that AI-generated content is automatically protected by "free speech." While art and parody are protected, these protections do not extend to fraud, impersonation, or the unauthorized sale of goods using a person’s identity. When AI is used to deceive an audience into believing a person said or did something they did not, that crosses the threshold from protected speech to illegal activity.
FAQ: Understanding AI and Legal Rights
1. Is it legal to make a deepfake of a celebrity if I label it as a parody? Parody is a legal defense, but it is not an absolute shield. Courts consider whether the audience is likely to be confused and whether the parody is meant to criticize the subject or simply to exploit their image for views or profit. If the content is indistinguishable from reality and causes professional or financial harm, parody claims often fail.
2. What should I do if I find an AI-generated video of me? Immediately report the content using the platform’s "Impersonation" or "Harassment" tools. If the platform does not act, gather documentation of the video, the URL, and any evidence of it causing you harm. Consult with an attorney specializing in digital media law to draft a formal takedown request or cease and desist notice.
3. Does the law protect my voice as well as my face? Yes. Increasingly, "voice clones" are being treated with the same legal weight as facial likenesses. Many jurisdictions are expanding their right-of-publicity statutes to explicitly include a person’s "vocal identity" to prevent AI from mimicking unique voice signatures for scams or unauthorized marketing.
4. Are AI companies liable for the things their users create? This is a developing area of law. While Section 230 of the Communications Decency Act (in the U.S.) has historically protected platforms from content created by users, courts are currently debating whether the tools themselves—the AI models designed to generate deepfakes—should carry liability if they are built with features that encourage or facilitate illegal impersonation.
5. How can I protect my digital identity from being used for deepfakes? Total prevention is difficult, but you can limit your risk by being cautious about the high-resolution images and audio clips you share publicly. Some professionals are now using "adversarial perturbations"—tiny, invisible changes to images that confuse AI models—to make their photos difficult for AI to scrape and train on successfully.
Recommendation: Protecting Your Future
The rise of generative AI necessitates a shift in how we approach our digital footprint. For creators, celebrities, and even private citizens, the most effective defense is a combination of proactive legal awareness and digital hygiene. Regularly monitor your digital presence using search alerts and reverse-image search tools to identify if your likeness is being misappropriated.
If you suspect you have been the subject of an unauthorized AI-generated deepfake, do not delay. Immediate action—documenting the source, reporting the violation to the platform, and, if necessary, seeking legal counsel—is the most effective way to protect your identity and prevent the spread of synthetic misinformation. As the law matures to address these challenges, staying informed about your rights is your most powerful tool in the digital age.
