The photography community is once again sounding the alarm over the legal language governing digital platforms, as VSCO finds itself at the center of a burgeoning controversy regarding its Terms of Use. The situation echoes a massive industry-wide backlash faced by Adobe nearly two years ago, when photographers discovered that the software giant’s updated terms appeared to grant the company overreaching access to user content. Now, VSCO, a platform long favored by professional and enthusiast photographers for its film-emulation filters and creator-centric community, is facing intense scrutiny over clauses that grant the company broad rights to user likenesses and the use of content for artificial intelligence training.
The controversy gained significant traction over the weekend after photographer Simon Migaj highlighted specific sections of VSCO’s Terms of Use on social media. Migaj pointed to language that requires users to grant VSCO the right to use the “name, image, voice, or likeness of any individual in your Content, in whole or in part, and in any form, media, or technology, whether now known or developed in the future.” Furthermore, the terms specify that VSCO maintains the right to use certain content to develop, train, and improve AI or machine learning models. While VSCO has published an updated set of terms scheduled to take effect on June 22, 2026, the core language regarding content licensing remains largely unchanged, suggesting that these policies have been in place, perhaps unnoticed by the broader user base, for some time.
The Evolution of the Terms of Use Controversy
The current friction between VSCO and its users is part of a broader historical trend in the software-as-a-service (SaaS) industry. In June 2024, Adobe experienced a public relations crisis when photographers noticed updated language in its Terms of Use that seemed to imply the company could access and use any work stored in the Creative Cloud to train its generative AI, Firefly. The backlash was so severe that Adobe was forced to issue multiple clarifications and eventually rewrite sections of its terms to explicitly state that it does not train AI on customer content stored locally or in the cloud.
The VSCO situation mirrors this trajectory but adds layers of complexity regarding the "right of publicity." By including "voice and likeness" in its licensing agreement, VSCO is entering a legal territory that goes beyond simple image hosting. This language is designed to protect the platform from lawsuits if a user’s face or voice appears in a featured post, a search result, or a promotional banner. However, in an era where "deepfakes" and AI-generated personas are becoming common, photographers are increasingly wary of any clause that cedes control over human identity to a corporate entity.
Understanding the "Scary" Legal Language
To the average user, the legal prose found in software agreements often reads like a total surrender of intellectual property. VSCO’s terms state that by using the service, creators grant a “royalty-free, sublicensable, non-exclusive, perpetual, irrevocable, worldwide license to use, reproduce, distribute, publicly perform, publicly display, and make derivative works” of their content.
While these terms sound predatory, legal experts and platform executives argue they are functionally necessary. For a platform to display an image on a smartphone, it must "reproduce" the file; to show it to other users, it must "distribute" and "publicly display" it. The "sublicensable" aspect allows the company to use third-party server providers, like Amazon Web Services or Google Cloud, to host and process the data. Without these broad permissions, a company like VSCO could technically be sued for copyright infringement simply for generating a thumbnail of a user’s photo.
VSCO CEO Eric Wittman addressed these concerns directly, emphasizing that the language is about operational viability rather than exploitation. "Without it, we couldn’t legally render your photos on different devices, let you share your profile publicly, or build the discovery features photographers rely on," Wittman stated. "This language protects our ability to operate, not to exploit."
The Likeness Clause and the Right of Publicity
The most contentious part of the current debate involves the "right of publicity." VSCO’s General Counsel, Sara Lee, explained that the inclusion of "name, image, voice, or likeness" is a standard legal safeguard. This clause ensures that if VSCO features a photographer’s work—perhaps a portrait or a video with audio—the company is not liable for violating the privacy or publicity rights of the person depicted in the content.
"We ask for permission to use these aspects of creators’ content so we can display and feature content and improve our platform," Lee said. She clarified that this does not equate to VSCO owning the work. The creator retains ownership, but the platform receives a license to use the attributes within the scope of its services. However, the phrase "whether now known or developed in the future" remains a sticking point for many. It is a "future-proofing" tactic that allows companies to adapt to new technologies—such as virtual reality or holographic displays—without needing to constantly re-negotiate terms with millions of users.
The Divide: Paying vs. Non-Paying Members in AI Training
Perhaps the most significant revelation in the current discourse is how VSCO differentiates between its subscription tiers regarding AI training. In a landscape where data is the new oil, VSCO has taken a bifurcated approach to how it feeds its machine learning models.
According to CEO Eric Wittman, VSCO does not use content from paying subscribers to train AI models or license it to third parties. This creates a "privacy-by-payment" model that is becoming increasingly common in the tech industry. For those who do not pay for the service, however, the rules are different. Publicly posted content from non-paying members is considered fair game for developing and improving AI-powered features.
"For non-paying members whose content is publicly posted, we may use that content to develop and improve AI-powered features on the platform," Wittman noted. He argued that this is a transparent acknowledgment of how most modern creative platforms operate. This data is used for features like "visually similar" recommendations, which rely on machine learning to analyze the composition and style of an image to suggest other relevant content to users.
Chronology of Platform Policy Shifts
The timeline of these changes reflects a rapid shift in the industry’s priorities toward AI integration:
- June 2024: Adobe faces massive backlash over TOS language, leading to a formal revision of their terms to protect user privacy regarding AI training.
- Late 2024 – Early 2025: Various social media platforms, including Meta (Instagram/Facebook) and Reddit, update their terms to clarify how user data is used for large language models (LLMs) and generative AI.
- May 2026: Photographer Simon Migaj flags VSCO’s "likeness" and "AI training" clauses, sparking a new wave of creator concern.
- June 2026: VSCO’s newly published Terms of Use are set to officially take effect, maintaining the controversial licensing language.
Industry Implications and the Trust Deficit
The recurring nature of these controversies points to a significant "trust deficit" between tech companies and the creative communities they serve. While the legal necessity of broad licensing is grounded in the reality of digital architecture, the lack of plain-language summaries often leads to a "gotcha" moment for users.
For photographers, the stakes are higher than for the average social media user. A professional photographer’s portfolio is their livelihood. The fear that a platform could use their unique style or the likeness of their models to train an AI that might eventually compete with them is a valid economic concern.
Furthermore, the "perpetual and irrevocable" nature of these licenses means that even if a user deletes their account, the platform may still retain the right to use the data it collected while the account was active, especially if that data has already been ingested into an AI training set. Removing data from a trained neural network is a notoriously difficult, and sometimes impossible, technical challenge.
Analysis: The Future of Creator Platforms
The VSCO controversy highlights a difficult choice for modern photographers. To reach an audience and use high-end editing tools, they must participate in ecosystems that demand significant legal concessions. The "opt-out" for AI training at VSCO is currently tied to a financial transaction—the subscription fee. This suggests a future where digital privacy and the protection of intellectual property become premium features rather than default rights.
As AI continues to evolve, the definition of "content" is expanding. It is no longer just the pixels in a JPEG; it is the metadata, the editing patterns, the facial structures, and the vocal frequencies. VSCO’s attempt at transparency—admitting that free users provide the data for AI development—is a step toward honesty, but it remains to be seen if the community will accept it as a fair trade.
For now, the advice for photographers remains consistent: read the fine print, understand the distinction between "ownership" and "licensing," and decide which platforms align with their professional ethics. The line in the sand is moving, and as VSCO has shown, even the most creator-friendly apps are not immune to the pressures of the AI revolution. The burden of vigilance continues to rest on the shoulders of the creators themselves, who must navigate an increasingly complex legal landscape to protect their work and their identities.
