Bria gives enterprises full control over how visual GenAI is trained, deployed, and governed, enforcing brand, data, and compliance rules at scale. Pulse 2.0 interviewed Bria founder and CEO Dr. Yair Adato to learn more.
Dr. Yair Adato’s Background
Could you tell me more about your background? Dr. Adato said:
“I’m the CEO and founder of Bria, where we’re building a world where human creativity and responsible AI innovation coexist and thrive together. My skill set spans many areas, but my secret power is bridging the gap between futuristic technologies, especially AI, and their commercial use and social impact. This ability to translate cutting-edge AI research into real-world applications that benefit society is what drives me every day.”
“I founded Bria in 2020 with a vision to create generative AI that respects artists, content creators and data ownership. That’s why we built our platform on 100% licensed data with our patented attribution engine – I believe technology should empower human creativity, not replace it.”
“Before starting Bria, I served as CTO at Trax Retail, where I played a pivotal role in scaling the company from a small startup with 20 employees to a unicorn with close to 1,000 staff members. I have a PhD in Computer Science in Computer Vision from Ben-Gurion University, in collaboration with Harvard University, and throughout my career, I’ve been granted more than 50 patents that build the bridge between AI innovation and commercial applications.”
“Scaling Trax from startup to unicorn taught me that AI only succeeds when it fits real production constraints: infrastructure, governance, and business accountability. That lesson shaped how we built Bria from day one.”
“On a personal note, I’m living in New Jersey with my wife and our beautiful 16-year-old twin daughters, and we have a cat who keeps us all entertained. Family time is really important for me – it keeps me grounded while we’re working on transforming how the world creates content.”
“My mission is simple: make AI work for people, not against them. Every line of code we write is guided by the principle that technology should amplify human creativity and respect data ownership, not replace it.”
Formation Of The Company
How did the idea for the company come together? Dr. Adato shared:
“When I first saw the first generative AI academia research in 2014, where Ian Goodfellow presented his work at a computer vision conference called ICCV, it immediately hit me in the stomach – I knew this was what I needed to do. It was clear to me that this technology would change everything we know about data creation, including visual data.”
“But then I realized that humanity’s prosperity is fundamentally built on IP and data ownership. That’s when I started seeking a solution for how data ownership could work in the era of generative AI. This isn’t just a technical challenge – it’s about preserving the economic foundation that allows creators and innovators to thrive.”
“This is where I met Vered Horesh, who shared the same vision. Interestingly enough, while I was thinking about this problem from the technology aspect, Vered perceived it from the legal perspective. Together, we realized we could create something revolutionary – a platform where generative AI and data ownership could coexist, where creators get compensated for their contributions to AI training, and where businesses could use powerful AI tools without legal risks.”
“Back in 2020, I met several true believers who supported the company – VCs with a futuristic view like Entree Capital and In Venture, and true partners like Getty Images who understood the vision from day one. These early supporters saw what we were building before the rest of the market understood the importance of responsible AI and data ownership.”
“We saw the writing on the wall early – generative AI would either become the next Napster, stealing from creators, or the next Spotify, fairly compensating them. We chose to build the Spotify model. Spotify succeeded not because it was ethical, but because licensing, economics, and rules were embedded directly into the infrastructure. Bria applies that same principle to visual AI for enterprises”
Favorite Memory
What has been your favorite memory working for the company so far? Dr. Adato reflected:
“Running a startup is a roller coaster – there are many happy moments as well as very hard ones. But one recent happy moment that really stands out is winning a Cannes Lions Gold award, a Clio, and a D&AD Pencil for our work together with Publicis – Marcel and Lidl for a user-generated content campaign.”
“This campaign presented such a cool implementation of generative AI in real life. It wasn’t just about the technology working – it was about seeing how our responsible AI platform could power a creative campaign that was both innovative and authentic, generating real engagement while maintaining brand safety and legal compliance.”
“What made it even more special was that this recognition came from the creative industry itself. These are some of the most prestigious awards in advertising and marketing, and seeing our technology being celebrated not just for its technical capabilities, but for enabling genuinely creative and commercially successful campaigns – that felt like validation of everything we’ve been working toward.”
“When Lidl’s campaign won a Cannes Lions using our technology, it proved that doing AI the right way doesn’t mean sacrificing creativity – it means unleashing it responsibly.”
Core Products
What are the company’s core products and features? Dr. Adato explained:
“Bria is built around three control layers that make GenAI production-ready:
- Control how AI behaves — Tailored generation, hyper-controllable models, and brand-locked parameters ensure consistent, predictable outputs.
- Control how AI is deployed — Multi-tenant cloud, bring-your-own-cloud, or fully on-prem, including air-gapped environments.
- Control how AI integrates — APIs, pipelines, source-available models, and iFrame apps that plug directly into existing workflows visual Generative AI
Within this platform, enterprises use modular capabilities which are all governed by the same control framework:
- Image Generation and Editing – We provide over 30 specialized APIs for creating and modifying visuals, from basic text-to-image generation to advanced editing capabilities like background generation, object removal, and image expansion. Think of it as having all of Adobe’s visual generation and editing capabilities as an API, all fully automated.
- Tailored Generation – This is our fine-tuning solution that allows companies to train models using their brand assets. It comes with 3 levels: zero-shot training, LoRA automatic training requiring only limited data, and lastly, deep training where hundreds or more of data is available. It ensures generated content aligns precisely with their visual identity, whether they need consistent brand colors, specific art styles, or character consistency.
- Product Shot Generation – Specifically designed for e-commerce and retail, this suite generates consistent product imagery (still and video) while maintaining product integrity. It’s perfect for creating lifestyle shots, product variations, and marketing materials at scale.
- Ads Generation – Our advertising and marketing automation tools that enable agencies and brands to create on-brand, personalized ads at scale while ensuring compliance and brand safety.
- Gen AI for Production Pipeline – We provide comprehensive tools and workflows that integrate seamlessly into existing production environments, enabling studios and creative teams to incorporate generative AI into their content creation pipelines efficiently.
All those products can be consumed as a Source-Available Models, API or application ready to be installed as an iFrame. For AI teams who need maximum control, we provide access to our foundation models’ source code and weights, including BRIA 3.2, our latest model that excels at text rendering and offers superior aesthetics in a compact 4-billion parameter architecture. For product and engineers teams we have the APIs capabilities.
What makes all of this unique is that everything is built on 100% licensed data with our patented attribution engine, which tracks content origins and ensures fair compensation for creators. We offer flexible deployment options – cloud, bring-your-own-cloud, or fully on-premises – ensuring complete data sovereignty and control.
From “Prompt and Pray” to Rules-Based Generation. A core differentiator behind Bria’s controllability is FIBO, a structured visual language that replaces black-box prompting. Instead of guessing and regenerating, teams can read, lock, and modify explicit visual parameters — preserving what works and changing only what they choose, without creative drift.
This enables deterministic, agent-ready workflows where AI becomes a reliable production system, not an experiment.
‘Think of us as the Visual Generative responsible AI infrastructure that enterprises can actually trust. We deliver enterprise-grade output quality with seamless integration into existing workflows – we’re not just another AI tool, we’re the legal, ethical foundation that makes AI safe for business.’”
Challenges Faced
Have you faced any challenges in your sector recently? Dr. Adato acknowledged:
“The biggest challenge we’ve faced is the fundamental tension between innovation and responsibility in generative AI. When we started Bria in 2020, most of the market (a very small one compared to today) focused purely on aesthetic output, but we saw that enterprises needed much more – legal certainty, brand consistency, and enterprise quality.”
“The challenge was that many companies were attracted to models trained on scraped internet data because they were free and produced impressive technology, but this is not exactly what they need. These models came with hidden risks – copyright violations, privacy infringement, regulatory compliance issues, and critically, the inability to copyright generated content.”
“We also had to redefine how quality is measured in generative AI. The industry was obsessed with making nice-looking images or video, but enterprise quality means consistency – getting the same brand colors and style every time; predictability – reliable results that meet brand guidelines; and customizability – fine-tuning models with specific brand assets so AI truly understands your visual language.”
“We overcame this by building the world’s largest repository of fully licensed training data through partnerships with Getty Images, Envato, and Alamy. We developed our patented attribution engine for creator compensation and content copyrightability, plus sophisticated fine-tuning for the consistency enterprises need.”
“The market is finally catching up to what we’ve known all along: pretty pictures aren’t enough. Enterprises need AI that’s predictable, controllable, and legally bulletproof.”
Evolution Of The Company’s Technology
How has the company’s technology evolved since its launch? Dr. Adato noted:
“Our technology evolution has been quite deliberate, building layer by layer to create a comprehensive visual gen AI platform as a service.”
“First, we solved the attribution technology and trained our text-to-image foundation models. We developed several foundation models for different tasks – for example, fast models for quick generation versus high-resolution models for detailed work.”
“Next, we developed predictable AI editing capabilities and product shot generation. This was crucial because enterprises need consistent, controllable results, not just creative outputs.”
“Then came our three-level tailored generation system. We offer few-shot learning, automatic LoRA fine-tuning with a few dozen images, or deep fine-tuning when hundreds of visuals exist. To explain these three levels, we like to give the example of customizing image generation: you might want something in the style of water lilies paintings, the style of Monet specifically, or the broader style of Impressionism.”
The next major step was wrapping all those capabilities into our Platform-as-a-Service. We made everything available at multiple levels – low-code interfaces for quick implementation, comprehensive APIs for developers, or full source code access for maximum control.
Now we’re expanding these same capabilities into video. We’re applying everything we’ve learned about responsible AI, attribution, fine-tuning, and enterprise-grade quality to motion content.
Significant Milestones
What have been some of the company’s most significant milestones? Dr. Adato cited:
“Some of our key milestones include launching our Visual Gen AI PaaS and securing strategic partnerships with leading enterprises and industry partners like Microsoft, AWS and NVIDIA.”
“Recently we raised funds to enable us to scale the platform and expand to new content types like music and enhance the video offering. Moreover, using these funds we’ve grown our team and technology infrastructure to support rapid adoption across industries, opening our main office in NY, with additional offices in LA and London.”
“We achieved over 300% growth in annual recurring revenue last year, showing the increasing demand for our solutions.”
Customer Success Stories
Can you share any specific customer success stories? Dr. Adato highlighted:
“Bria achieved tremendous growth among new enterprise customers over the past year, partnering with global brands in media and entertainment, retail, and creative platforms.”
“We are the generative AI generation and AI editing technology behind Getty Images and iStock, powering their visual creation capabilities for millions of users worldwide. We’re also the AI editing engine in Envato by Shutterstock, enabling their creative platform to offer advanced generative capabilities to designers and content creators globally.”
“A notable collaboration with Publicis and MARCEL for Lidl utilized Bria’s AI-driven foundation models for a personalized UGC campaign called ‘Lidlize.’ This campaign became a viral sensation, allowing users to transform everyday items into Lidl’s distinctive visual style. The campaign won multiple prestigious awards including Cannes Lions Gold, Clio, and D&AD Pencil. The AI engine was deployed as a brand-owned system that enforced Lidl’s rules, not a generic model’s defaults.”
“Another significant partnership is with major CPG companies where our tailored generation capabilities enable them to create on-brand content at scale while maintaining strict brand guidelines. Our platform allows their creative teams to generate consistent, high-quality marketing materials in minutes rather than weeks.”
Funding/Revenue
Are you able to discuss funding and/or revenue metrics? Dr. Adato revealed:
“Yes, in our Series B funding round, we raised $40 million, bringing our total capital raised to $65 million. Our revenue has grown substantially, with our annual recurring revenue increasing by more than 300% last year. We’ve been able to scale quickly due to our enterprise-focused model and the efficiency of our platform.”
Total Addressable Market (TAM)
What total addressable market (TAM) size is the company pursuing? Dr. Adato assessed:
“We’re targeting a vast market, with our solutions applicable across industries like marketing, e-commerce, retail, media, gaming, and entertainment. The TAM for generative AI is growing rapidly as more enterprises seek scalable, compliant content creation tools. Our focus is on providing enterprise-grade solutions that deliver measurable impact, which positions us well in this expanding market.”
Differentiation From The Competition
What differentiates the company from its competition? Dr. Adato affirmed:
“Bria is built for enterprises that need GenAI to operate inside real business constraints. Your. AI. Your Rules means customers define deployment, data boundaries, brand consistency, and IP protection; and the platform enforces those rules by design. Most tools generate images. Bria delivers governed, production-grade outcomes that survive legal, brand, and compliance review.”
“We lead with IP protection, deliver creativity that’s predictable and customizable, and offer PaaS flexibility that meets customers where they are.”
Future Company Goals
What are some of the company’s future goals? Dr. Adato emphasized:
“We’re currently adding video capabilities to our visual generative AI PaaS. This is a natural evolution of our platform – taking everything we’ve learned about responsible AI, attribution, fine-tuning, and enterprise-grade quality from images and extending it to motion content.”
“A major focus area is what we call ‘Premium IP’ – we’re pioneering an innovative model for premium content or “copyright streaming” in visual media. This enables IP holders, from major studios to individual creators, to monetize their assets through our platform via customized pricing tiers based on usage rights. Think of it like how Spotify transformed music distribution.”
“This Premium IP initiative could unlock entirely new revenue streams for content owners while providing our customers with access to high-value, properly licensed assets. A Marvel Studios executive, for example, could offer their Spider-Man IP through the platform at different price tiers: $50 for social media marketers, $500 for merchandise designers, and $5,000 for film production teams.”
“We’re reimagining how the entire creative economy can thrive in the age of AI. Premium IP is about turning AI from a threat to creators into their biggest revenue opportunity.”
Additional Thoughts
Any other topics you would like to discuss? Dr. Adato concluded:
“I think it’s crucial to continue the conversation about the ethical and legal implications of generative AI. At Bria, we fundamentally believe that Gen AI should empower people, not threaten them.”
“Bria’s commitment to responsible AI begins with our foundation in exclusively licensed data, training models from scratch using partnerships with premium providers like Getty Images, Envato, Alamy, Deposit Photos, and Freepik. This approach ensures legal compliance, prevents privacy and copyright infringement, and delivers superior quality outputs that enterprises can confidently deploy.”
“We employ a sophisticated strategy to address bias and fairness: combining diverse, high-quality data repositories from multiple sources with proprietary algorithms applied during both training and inference stages. This comprehensive methodology has positioned Bria as an industry leader in producing fair and balanced visual AI-generated content.”
“Bria takes a proactive approach to safety. There are no images of famous people in our training data, effectively preventing the creation of potentially harmful or misleading deepfakes. This preventative measure, combined with our multi-layered bias mitigation techniques, creates an AI platform that enterprises can trust to minimize reputational risks while delivering powerful generative capabilities.”
“Overall, we’re committed to ensuring that AI is used responsibly, and that both creators and enterprises are protected. We’re focused on creating products that empower businesses to innovate without compromising on their values.”
“The future belongs to AI companies that empower rather than exploit. We’re building that future, one licensed pixel at a time.”
