Preliminary Injunction Issued in the Sora “Cameo” Trademark Dispute: What It Signals for AI Product Strategy

by Jay Kotzker

On February 14, 2026, the United States District Court for the Northern District of California issued an Order Granting Motion for Preliminary Injunction in Baron App, Inc. d/b/a Cameo v. OpenAI, Inc., et al., Case No. 25-cv-09268-EKL . The decision follows a temporary restraining order entered in November 2025 and materially alters the litigation landscape. The court has now determined that the plaintiff, Baron App—the operator of the CAMEO® platform—is likely to succeed on the merits of its trademark infringement claim.

For AI companies developing consumer-facing products, particularly those operating in entertainment-adjacent markets, the ruling carries implications well beyond a naming dispute. It signals how courts are applying established trademark doctrine to feature-level branding within emerging AI ecosystems.

What the Court Decided

The court granted a preliminary injunction enjoining OpenAI from using the federally registered CAMEO® mark in connection with its Sora application. The court found that Cameo owns valid federal trademark registrations and that OpenAI used the identical term “cameo” in connection with Sora 2. It further concluded that Cameo demonstrated a likelihood of success on the merits of its infringement claim, that irreparable harm was likely in the absence of injunctive relief, and that the balance of equities and public interest favored an injunction.

OpenAI had already renamed the feature following the TRO. The preliminary injunction extends that restriction through the pendency of the litigation. This is no longer a temporary procedural safeguard; it is a substantive judicial finding that the plaintiff is likely to prevail.

How the Court Reached Its Conclusion

The court’s analysis centered on the Ninth Circuit’s eight-factor Sleekcraft test for likelihood of confusion. Of the eight factors, five weighed in Cameo’s favor and three were neutral. None favored OpenAI.

With respect to the strength of the mark, the court acknowledged that “Cameo” may be conceptually suggestive or descriptive. However, it emphasized the mark’s commercial strength, pointing to extensive media coverage, millions of social media followers, viral distribution, and marketplace recognition . The decision underscores a key principle in trademark law: even a linguistically ordinary term can become legally powerful when market recognition is substantial.

The relatedness of the services proved especially significant. Both platforms offered vertically formatted, shareable, personalized celebrity-style videos. The court noted that certain AI-generated Sora videos were visually indistinguishable from authentic Cameo videos . The fact that one product involved AI-generated content and the other human-recorded content did not eliminate competitive proximity. In trademark analysis, technological differences do not negate relatedness if consumer perception overlaps.

Similarity of the marks was straightforward. The term at issue was identical in sight, sound, and meaning. The court treated this factor as weighing heavily in Cameo’s favor . The analysis did not hinge on corporate branding but on feature-level identity within the product.

The court also credited evidence of actual confusion. An Eveready survey reflected a net confusion rate of 10 percent, which courts often treat as probative when supported by additional evidence . The record included misdirected customer inquiries, social media mis-tags, Reddit commentary, and instances in which consumers questioned whether the companies had partnered. Taken together, the court found that this evidence supported a likelihood of confusion.

Notably, the court framed the issue primarily as one of forward confusion—the risk that consumers would believe Sora’s feature was affiliated with, endorsed by, or sponsored by Cameo . In modern AI markets, where cross-platform partnerships and licensing arrangements are common, courts may view affiliation confusion as particularly plausible.

OpenAI asserted a descriptive fair use defense. To prevail, a defendant must show that the term is not used as a trademark, that it is used descriptively, and that it is used in good faith. The court concluded that OpenAI had not demonstrated a likelihood of success on this defense at the preliminary injunction stage . The ruling reinforces that invoking a dictionary meaning is insufficient where a term is used prominently as a feature name and marketing identifier.

Why This Matters for AI Companies

The decision is not about the legitimacy of generative video technology. It is about brand architecture in emerging technology markets.

First, the ruling illustrates that feature names carry independent legal risk. The court’s analysis focused not on the defendant’s corporate name but on the branding of a specific feature within the product. In AI platforms, features frequently become the focal point of user engagement, marketing narratives, and viral distribution. When a feature name functions as a source identifier, it will be evaluated under trademark law accordingly.

Second, AI tools often operate adjacent to crowded and highly developed brand ecosystems, including entertainment, creator marketplaces, advertising platforms, and media infrastructure. Even if the underlying technology differs, adjacency can heighten the likelihood of confusion analysis. The court made clear that technological novelty does not insulate branding decisions from traditional trademark scrutiny.

Third, the procedural timeline demonstrates how quickly courts may intervene. This dispute moved from product launch to TRO to preliminary injunction within months . For high-growth companies, that timeline can compress rebranding costs, product messaging shifts, user acquisition strategies, and investor communications. Trademark litigation in AI markets is no longer a slow-burn risk; it can become operationally immediate.

The Strategic Layer: Innovation Speed vs. IP Maturity

AI product teams often iterate faster than conventional brand clearance workflows. Yet courts continue to apply established trademark doctrine without adjustment for technological novelty. The Sora/Cameo dispute demonstrates that identical marks, related digital video services, and overlapping marketing channels can result in injunctive relief—even in cutting-edge AI contexts.

Innovation velocity does not diminish the importance of IP diligence. Legal architecture must evolve at the same pace as product architecture.

Holon’s Perspective

At Holon Law Partners, we approach emerging technology disputes not as isolated conflicts but as signals of structural change. This ruling underscores a central reality: in high-velocity and regulated sectors, trademark strategy is not cosmetic. It is a component of operational risk management.

Companies building in AI, cannabis, biotech, and other emerging industries should approach feature naming and brand development with the same rigor applied to product design. Early clearance, ecosystem mapping, and forward-looking confusion analysis are no longer optional safeguards—they are strategic imperatives.

Once a preliminary injunction is issued, leverage has already shifted.

This article is provided for informational purposes only and does not constitute legal advice. It does not create an attorney-client relationship. Organizations should consult qualified counsel regarding their specific circumstances.