Getty Images, WPP, and Publicis Are Already Using Bria's Licensed AI Models

You start with a prompt, something safe and corporate: "a smiling woman in a blue blazer, holding a tablet, modern office background." The image generates, crisp and professional. The real work begins a click later, in the fine print. You scroll past the model card, past the API call, to the provenance tab. There, Bria lists the stock photo IDs,Getty 54321, Alamy 98765,that influenced the pixels on screen, each one licensed, each one traceable. For a brand lawyer, this isn't a feature; it's the product.

Bria, founded in Tel Aviv in 2020, sells visual AI that doesn't apologize. Its models are trained from scratch on data licensed from Getty Images, Shutterstock, Alamy, and Envato, a curated library paid for by the byte. The output is an enterprise-grade image or video, but the real deliverable is a legal receipt. In a year when copyright lawsuits have become a standard cost of doing business for AI companies, Bria's pitch is a simple invoice. It charges $0.03 per generated image and, in return, offers indemnification. For the marketing teams at holding companies like WPP and Publicis Groupe, which are listed as customers, that trade is the entire point [Intel Capital, Unknown] [Perplexity Sonar Pro Brief, Unknown].

The wedge is the waiver

Bria's competition isn't just other image generators; it's corporate risk departments. General-purpose models, for all their creative horsepower, operate in a legal gray area built on scraped web data. Bria's foundation is a different asset: exclusive licensing deals with major stock providers and a patented attribution system that tracks how specific licensed photos influence outputs, funneling compensation back to the original creators [Perplexity Sonar Pro Brief, Unknown]. This isn't an AI wrapper on top of Stable Diffusion. The company says it trains its own foundation models, and it has released an open-source text-to-image model claimed to match leading performance with roughly one-third the parameters, a nod to efficiency for developers [Bria, Unknown].

The bet is that controllability and compliance will matter more than raw capability for the enterprise budgets that matter most. The platform emphasizes fine-grained control over brand consistency,exact colors, specific layouts, defined art styles,to slot into existing production workflows. The customer list suggests the wedge is working. Beyond the agency giants, tech platforms like Getty Images itself and Imgix use Bria to power their own visual AI features, a powerful endorsement from the very rights-holders it partners with [Bria, Unknown].

Founders from the data trenches

The company is led by co-founders Dr. Yair Adato and Gal Jacobi. Adato, the CEO, is a computer vision PhD whose most recent operational role was VP of Data Science and Machine Learning at Fiverr, the freelancer marketplace [Microsoft, Unknown]. That experience likely ingrained the complexities of commercial content licensing at scale. Jacobi, a serial entrepreneur, brings a product and business lens from past roles in AI and data-oriented companies [AI Market Watch, May 2026]. Their backgrounds point to a company built not just on research papers but on the practical mechanics of monetizing digital creativity.

Traction has attracted capital. Bria has raised a total of $65 million, with a $40 million Series B in 2025 led by Red Dot Capital Partners [PRNewswire, March 2025]. The investor syndicate includes strategic names like Intel Capital and advertising giant Publicis Groupe, which is both an investor and a customer.

2024 Series A | 24 | M USD
2025 Series B | 40 | M USD
Total Raised | 65 | M USD

The risks of a clean room

For all its legal armor, Bria faces distinct challenges. Its model is a trade-off: licensed data ensures safety but may limit the creative diversity and cultural zeitgeist capture of models trained on the entire internet. The platform's success hinges on enterprises prioritizing risk mitigation over the bleeding edge of AI artistry.

• The capability gap. If general-purpose models achieve legal clarity through industry-wide licensing deals or court precedents, Bria's primary differentiator could erode. Its answer is to compete on controllability and workflow integration, not just the training dataset.
• The cost of cleanliness. Licensing data is expensive, a cost presumably baked into Bria's pricing. The company must prove its value exceeds the premium, convincing customers that its $0.03 per image is cheaper than a single copyright claim.
• Market education. Bria must sell the abstract concept of "reduced legal risk" in a market accustomed to evaluating tools on output quality and speed alone. Its partnerships with stock agencies and public case studies, like a Nutella campaign run on AWS infrastructure, are key to this translation [AWS, Unknown].

The next twelve months

The roadmap is about deepening the enterprise embrace. Expect more named customer launches from the Fortune 500, particularly in regulated sectors like pharmaceuticals and finance where compliance is non-negotiable. The company's hybrid headquarters in New York and Tel Aviv support a global sales push. Technically, the focus will likely be on expanding its video generation capabilities and refining the developer experience around its APIs and source-available models.

The cultural question Bria is answering isn't about whether AI can make art. It's about whether corporations can afford to let it. In a boardroom, the most beautiful AI-generated image is worthless if it comes with a subpoena. Bria is betting that for the global brands shaping our visual environment, safety will always be the most creative choice.

Sources

1. [Intel Capital, Unknown] BRIA Raises $24M Series A to Pioneer Responsible Visual Generative AI For Enterprises | https://www.intelcapital.com/bria-raises-24m-series-a-to-pioneer-responsible-visual-generative-ai-for-enterprises/
2. [Perplexity Sonar Pro Brief, Unknown] Bria.ai company overview
3. [Bria, Unknown] Models | https://bria.ai/models
4. [Microsoft, Unknown] Microsoft for Startups feature
5. [AI Market Watch, May 2026] Bria AI profile
6. [PRNewswire, March 2025] Bria Secures $40M Series B | https://www.prnewswire.com/news-releases/bria-secures-40m-series-b-302087123.html
7. [AWS, Unknown] AWS case study: Bria AI and Nutella campaign
8. [TechCrunch, March 2025] Bria lands new funding for AI models trained on licensed data | https://techcrunch.com/2025/03/13/bria-lands-new-funding-for-ai-models-trained-on-licensed-data/