Adobe used his photo library to train AI — now Gerald Carter is protecting music producers with Pre-Label

For nearly a decade, Gerald Carter worked to fill a gap in the stock photo industry with his company, Diversity Photos. After seeing how difficult it was for his wife, a marketing executive, to find images that fully reflected Black life, he built a digital library rooted in representation. What began as a personal frustration grew into a business centered on visibility and ownership.

Now, Carter is turning his attention to music.

His latest venture, Pre-Label, is an AI platform designed to reshape how Black producers are protected and paid as artificial intelligence continues to transform the industry. The platform allows producers to convert their existing catalogs into owned, monetizable assets, rather than relying on labels or distributors to define the terms.

Lessons From the Tech Industry

Carter’s approach is shaped by his experience working with Adobe.

After launching Diversity Photos in 2016, Carter and his team invested heavily in building the platform, collaborating with more than 100 photographers and hundreds of subjects. The company eventually secured licensing deals with major players, including Adobe, in what was initially framed as a partnership.

That relationship shifted in 2023 when Adobe introduced Firefly, its generative AI image tool. Carter discovered that images from the Diversity Photos catalog had been used to train AI systems without his knowledge or additional compensation.

Adobe offered $1,100 for the use of nearly 12,000 images. For Carter, the offer underscored how little control creators can have once their work enters large platforms.

Although his contract provided some protection, the cost of legal action forced the case to be dismissed before his claims could be fully examined.

Building Ownership from the Start

The experience reshaped Carter’s perspective.

Rather than waiting for platforms to establish fair terms, he saw the need to build ownership into the foundation. That thinking led to Pre-Label.

For generations, Black creativity has driven cultural and commercial value, from blues and jazz to hip-hop, while ownership has often remained out of reach. As AI advances, the ability for systems to absorb and replicate entire catalogs raises new concerns, particularly for Black creatives.

How Pre-Label Works

Carter co-founded Pre-Label with producer Pharren Lowther, whose credits include work with Future and Lil Wayne. Together, they aim to close the gap between cultural impact and financial return.

Through the platform, producers can upload their catalogs, which are then organized and used to train a personal AI model known as a “Producer Twin.” Built entirely from a producer’s own work, the model allows fans and artists to generate beats in that producer’s style.

Producers retain ownership, set their own pricing and collect revenue directly, avoiding traditional splits with labels or distributors.

Reclaiming Control in the Age of AI

Pre-Label reimagines the same technology often used to extract value from creative labor, instead positioning it as a tool for protection and control.

The distinction is critical. Compensation after the fact is not the same as ownership from the beginning. By centering control at the infrastructure level, Pre-Label offers a different path forward for creators navigating an AI-driven industry.

The Black Wall Street Times spoke with Gerald Carter about Pre-Label’s mission to help Black creatives use AI to own and protect their work.

The Black Wall Street Times: You went from fighting a system that extracted value from your work to building a platform designed to let creators capture that value from the start. What was the moment, or the realization, that pushed you toward building Pre-Label?

Gerald Carter: The realization came when I understood that most creators are taught to focus on visibility before ownership. By the time many artists, producers, photographers, or writers realize their work has real value, the infrastructure around that work is already owned by someone else, whether that’s a platform, a distributor, or now increasingly, an AI system.

What happened with my own work made that impossible to ignore. I spent years building Diversity Photos to fill a representation gap that traditional stock platforms ignored. I’m still scaling that company. That archive had cultural value because it reflected real people, real communities, and authentic experiences that were historically absent from digital media. Watching the broader AI industry aggressively pursue datasets like that without meaningful conversations around consent, compensation, or long-term ownership clarified something for me: creators cannot afford to think about ownership later anymore.

Pre-Label was born from the idea that ownership has to begin at the creation stage itself. Not after the deal. Not after the upload. Not after the exploitation has already happened.

Pre-Label’s core idea is that creators need to establish ownership at the infrastructure level, not after the fact. What does that actually look like in practice for a music producer today? What does Pre-Label give them that they don’t currently have, and how does it teach them to obtain it?

GC: For producers today, ownership is often fragmented before they even understand what they own. They may create a beat, send files through multiple platforms, collaborate informally through text messages or DMs, and never establish proper metadata, registration, licensing structure, or usage terms upfront. Then years later, they’re chasing royalties or trying to prove authorship. Infrastructure-level ownership means embedding protection, documentation, and monetization into the creative workflow itself.

Pre-Label is designed to help producers establish verifiable ownership from day one. That includes data provenance; maintaining control over licensing; educating creators about publishing rights, splits, metadata, and AI training implications and opportunities; and helping them understand that intellectual property is an asset class, not just content. Producers use their own IP to create producer twins that integrate into their current creative workflow – assisting (not replacing) the producer.

A lot of creators have talent. What they haven’t historically had is access to the legal, technological, and business infrastructure that larger corporations use to protect and scale value. Pre-Label aims to close that gap.

Super producer Pharren Lowther helped define the sound of some of the industry’s biggest artists, yet saw limited direct financial return from those streams. How do you think that experience shaped the mission behind Pre-Label?

Pharren’s experience reflects a larger reality across the music industry: cultural contribution and financial participation are often disconnected, especially for Black creators operating behind the scenes.

You can shape the sound of an era and still not fully participate in the long-term value generated from your work. That’s not because the contribution lacked value. It’s because the systems around ownership, publishing, licensing, and rights management were historically built in ways that creators themselves often weren’t fully included in.

That reality deeply influenced Pre-Label’s mission. We’re trying to help creators think beyond creation and toward leverage. Beyond virality and toward equity. Beyond placement and toward ownership. The goal is not simply helping creators “get paid.” It’s helping them build durable assets that continue generating value over time.

Diversity Photos, was built with digital autonomy that provided images of real people, real diversity, real consent. Did you think about that archive not just as a business, but as a kind of cultural responsibility? Did you ever anticipate it would become part of the early works of AI?

GC: From the beginning, Diversity Photos was about correcting a distortion. Digital imagery shapes public perception, and for a long time Black communities and other marginalized groups were either absent from stock photography or represented through stereotypes.

I viewed the archive as documentation as much as commerce. These were real people consenting to be represented authentically in digital spaces where authenticity historically did not exist.

At the time, I absolutely understood the archive had long-term technological relevance, but I don’t think most people fully anticipated how aggressively AI systems would rely on existing image repositories to build generative capabilities. What became clear very quickly is that archives are no longer passive collections. They are fuel for machine learning systems.

That changes the ethical conversation entirely.

Diversity Photos was built specifically because Black and minority communities have been historically underrepresented in stock photography. Now that archive was used to train an AI that can generate that imagery on demand. How do you feel about that?

GC: There’s a complicated emotional reality there. On one hand, it further validates the importance of the archive. The reason these systems sought out diverse datasets is that authentic representation mattered and was missing.

But on the other hand, it sucks seeing communities that were historically excluded suddenly become valuable only once their likeness, culture, and identity became commercially useful for AI systems. That tension is real. Representation without ownership is exploitation. The issue is not simply whether Black communities are included in AI. The question is whether they participate in the economic value being created from that inclusion.

You turned down involvement with Shutterstock in 2021 when they wanted to use Diversity Photos to train their AI. What was your reasoning at the time, and how do you look back on that decision now?

GC: At the time, my concern was simple: the industry had not yet developed meaningful frameworks around consent, compensation, attribution, or downstream ownership regarding AI training data. Once you contribute to a training ecosystem, you often lose visibility into how that data is being used, replicated, commercialized, or redistributed. I was uncomfortable participating in that without stronger protections in place.

Looking back, I still believe it was the right decision. Not because AI itself is inherently negative, but because creators, especially historically marginalized creators, deserved more voice in those conversations than they were being given.

We are now watching entire industries debate issues many creators raised years ago.

There’s a long pattern of Black cultural production being absorbed into mainstream systems without the originators seeing proportional return, especially within the digital space. Is AI a new approach to that silencing, or the same chapter with better tools?

GC: I think it’s an acceleration of an existing pattern. Black culture has consistently functioned as a source material economy for larger systems, from music to fashion to language to digital trends. The difference now is speed and scale. AI can absorb, replicate, remix, and monetize cultural expression at unprecedented velocity. That creates enormous risks if ownership structures remain unequal.

At the same time, I think AI also presents an opportunity. Historically, gatekeepers controlled distribution. AI potentially lowers barriers to creation and scalability for independent creators. But that opportunity only matters if creators maintain ownership over their data, likeness, and intellectual property. Otherwise, we repeat the same extraction cycle under more sophisticated branding.

Research shows that AI perpetuates harmful racial stereotypes while simultaneously extracting and commodifying authentic Black cultural expression without providing any benefit. How is Pre-Label leading the charge to where Black voices not only are represented in the data utilized by AI, but now own their own data?

GC: Ownership is the central issue. For years, the conversation around diversity in technology focused primarily on inclusion. But inclusion without governance does not create equity.

Pre-Label is focused on helping creators establish systems where their intellectual property, metadata, creative outputs, and training contributions remain connected to them as assets. That means building infrastructure around provenance, licensing, verification, and monetization rather than simply encouraging creators to upload content into systems they don’t control. We also believe education matters. Many creators still don’t fully understand how AI systems ingest and learn from data. Once creators understand that their work is training material, they begin thinking differently about contracts, uploads, permissions, and platform participation.

The future conversation cannot just be “Are Black creators visible?” It has to become “Do Black creators own meaningful portions of the value chain?”

What do you think creators across any medium can do to protect their IP from the continued intrusion of AI? How is Pre-Label providing users with that security?

GC: Creators need to begin thinking like rights holders, not just creatives. That means documenting authorship, registering work properly, maintaining organized metadata, understanding licensing terms, reading platform agreements carefully, embedding watermarks, and becoming more intentional about where and how content is distributed.

A lot of creators unknowingly sign away rights simply because the systems are designed for speed and convenience over protection. Pre-Label’s role is to simplify that process and build tools that allow creators to secure ownership earlier and more transparently. The goal is not fear-based resistance to AI. AI is here. The question is whether creators enter this next era positioned as owners or merely as raw material suppliers.

You have been at the forefront of AI before it became a basic consumer product. With your firsthand experience, how do you see AI including Black voices in the future? And how can Black creators do their best to be part of the continued implementation?

GC: I think Black voices will absolutely shape the future of AI because Black culture has historically shaped almost every major shift in digital culture. The real question is whether Black creators will participate as stakeholders or simply as contributors.

That requires moving beyond consumption into building, ownership, policy, and technical literacy. Creators should not only learn how to use AI tools. They should understand how datasets work, how licensing works, how intellectual property functions inside machine learning ecosystems, and how value is created at the platform level.

The creators who thrive in this next era will be the ones who combine creativity with ownership strategy.

That’s ultimately what Pre-Label is about: helping creators build futures where their cultural contributions remain connected to long-term economic power.

The post Adobe Used His Photo Library To Train AI — Now Gerald Carter Is Protecting Music Producers With Pre-Label appeared first on The Black Wall Street Times.