BRIA lands new funds for AI models in license data

Image generators with AI, who are in the center of a series of author rights demands against AI companies, are frequently trained in massive amounts of public websites data. Most of these companies argue that Fair use doctrine Protects its scratch and data training practices. But many copyright holders do not agree.

That is why some new companies and companies that develop image generators are testing a different tactic: training generators exclusively in license content. Based in New York and Tel Aviv BriaFounded in 2023 by businessmen who Yair Adato and Assa Eldar, is one of these.

Bria pays for images of around 20 partners, including Getty Images, and uses them to train image generation models with content railings. Adato, the CEO of Bria, said the “programmatically” platform compensates for image owners according to their “general influence.”

“The Bria Foundation models house one billion images and millions of videos,” Adato told Techcruch. “Bria has mitigated biases that can sometimes arise in the visual content generated by training their models in globally representative data sets. The company’s models constantly produce images that reflect diversity, making them suitable for several creative applications. “

Bria offers accessories for image editing and design applications that include Photoshop and Figma, as well as an adjusted API that allows customers to customize the company’s models for specific applications. Users can execute Bria models on the company’s platform or in an external computer environment, such as a public cloud. In any case, customers have the data and exits, said Adato.

“Business customers can pay access to the source code and (models),” said Adato. “We provide more than 30 specialized APIs to create and modify images, which customers access through the subscription and prices based on use. Companies can pay to adjust our generative AI models with their brand assets, creating custom engines that maintain their visual identity. “

Bria’s plans are ambitious. Adato tells TechCrunch that the company of 40 people seeks to promote an “IP ecosystem” where companies can access images with a media conglomerates for use in commercial creations, with “incorporated compliance.”

Bria also plans to expand its platform and models to admit additional types of media, including music, video and text, as well as applications on the device.

“Bria continues to thrive despite the widest challenges of the technology industry,” said Adato. “While the sector faces winds against the maturation of the technology company market, macroeconomic pressures that cause budgetary limitations and the oversaturation of basic AI wrapping applications, these factors strengthen Bria’s position.”

While a growing number of companies try to build businesses around licensed media generators, including Adobe, Spawning AI and Shuttersock, Bria has managed to obtain a support point in the nascent market. On Thursday, the company announced that it raised $ 40 million in a round of financing of the B series directed by Red Dot Capital with the participation of Maor Investment, Entée Capital, Gft Ventures, Intel Capital and In-Venture.

After Bria’s total up to around $ 65 million, most of the new cash will be allocated to the development of products, said Adato.

“We are growing rapidly with our 40 clients, demonstrating a significant annual growth of recurrent income of more than 400% last year,” said Adato. “We are also expanding to our team with additional experience in several key areas: researchers and generative engineers of AI in music and videos, global sales and marketing leaders, IP experts and copyright, and generative consultants of AI. We hope to double the size of our team by the end of the year. “