While big tech giants like Apple and Alphabet are making content generation a cornerstone of their respective AI initiatives, startups are raising record sums to bring AI-generated visuals to enterprise clients.
Unlike large language models, which take a jack-of-all-trades approach, these software startups solve specific business needs—from helping retailers edit product shots to allowing sales teams to create personalized videos for prospects.
In Q1, deal value for AI visual media, a subsegment of generative AI, increased nearly 90% to $402.7 million from the previous quarter’s total of $212.7 million, according to PitchBook data. It’s the highest quarterly total in recent years. Deal value spiked in mid-2023 when enterprise-focused Runway, an AI video editing and generation startup, raised a $141 million Series C extension that June backed by Nvidia, Salesforce Ventures, and Google.
Enterprise use cases are powering much of the momentum, particularly in marketing and design practices.
Tavus, a generative video startup for sales and outreach training, raised an $18 million Series A led by Scale Venture Partners in March. Photoroom, which offers an AI photo editor aimed at ecommerce businesses, raised a $43 million Series B in February led by Balderton Capital, Aglaé and Y Combinator. And Kittl, the developer of an AI asset generation tool for graphic designers, raised a $36 million Series B led by IVP in January.
Meanwhile, big tech’s push into the field threatens consumer-facing startups that rode a wave of adoption for the technology in recent years.
Apple recently unveiled Apple Intelligence, a localized generative AI assistant capable of creating unique emojis, art and images of personal contacts. At Google’s I/O presentation in May, it unveiled Veo, a video generation application running on a specialized model, as well asImagen 3, a text-to-image generator.
Consumer-focused AI media generation startups have recently faced headwinds as competition proliferates. Stability AI, the creator of the text-to-image model generator Stable Diffusion, has reportedly been exploring a sale as it faces a cash crunch following layoffs and the resignation of CEO Emad Mostaque.
Legal concerns have also plagued the space. Stability and Midjourney, another popular text-to-image creator, have become embroiled in lawsuits from artists who claim the startups used their work to train their respective models. Both companies deny wrongdoing.
While some are concerned about what big tech’s entry into the space could mean for their bottom line, other content-generation startups have welcomed the competition.
“For us, this is good news,” said Yair Adato, CEO and founder of Bria, an image generation startup geared toward advertising and other commercial use cases. Companies like Apple are increasing awareness of the potential for AI media generation, which helps companies like Bria, Adato said. The startup recently raised a $24 million Series A in February from investors like Intel Capital and Samsung Next.
Adato said that startups have an important role to play in keeping the AI ecosystem open and accessible. Big tech companies having a monopoly on AI models and technology, Adato said, cannot go unchallenged.
“The day when enterprise and developer teams do not have access to the source code of foundation models will be the day innovation dies,” he said.
Related read: Artificial Intelligence & Machine Learning Report
Featured image by Qi Yang/Getty Images
Learn more about our editorial standards.
