Adobe Introduces Firefly Foundry to Help Brands Build Private Generative AI Models

At its annual Adobe MAX conference, Adobe announced the launch of Adobe Firefly Foundry, a service designed to help businesses create proprietary generative AI models trained on their own intellectual property. The company describes the initiative as a way for brands to develop AI tools that reflect their visual identity while maintaining control over data and creative output.

The new offering builds on Adobe’s existing Firefly platform, which provides generative AI tools for creating images, videos, audio, and 3D assets. Firefly Foundry extends that capability by allowing organizations to work directly with Adobe to produce custom AI models securely trained on their own branded content. The service integrates with Adobe’s broader ecosystem, including Creative Cloud, Express, and GenStudio, allowing businesses to generate on-brand materials at scale.

Hannah Elsakr, vice president of GenAI New Business Ventures at Adobe, said the Foundry is aimed at solving “complex content and media production challenges.” She added that Adobe has already collaborated with companies such as Walt Disney Imagineering to explore how the technology can deepen audience engagement.

Responding to Soaring Content Demands

Adobe’s move comes amid growing demand for on-brand digital content. A company study cited in the announcement suggests marketers expect content needs to rise more than fivefold within two years. For many large organizations with extensive product portfolios and long-established visual identities, maintaining consistent brand presentation across thousands of assets has become a significant operational challenge.

Generative AI is being positioned as a potential solution, capable of accelerating content creation while maintaining brand standards. Firefly Foundry aims to meet that demand by giving marketing teams access to AI models built on their existing design archives. The models can generate new imagery, video, and other assets that align with brand aesthetics, potentially reducing production costs and turnaround times.

Focus on Control and Safety

Adobe has consistently promoted its AI products as being “commercially safe,” meaning they are trained on licensed or Adobe-owned data rather than scraped internet content. This has become a key differentiator as concerns grow over copyright and data privacy in generative AI. Firefly Foundry continues that positioning, offering what Adobe calls a “responsible AI” framework.

Businesses using the service will be able to test AI outputs, manage model access, and deploy their custom models across internal teams. Adobe also said it will provide embedded specialists—including AI researchers and engineers—to collaborate with clients on tailored use cases.

Competitive and Industry Context

The launch of Firefly Foundry comes as enterprise adoption of generative AI accelerates. Technology providers such as OpenAI, Google, and Anthropic are courting large organizations with customizable AI solutions that can integrate brand data or proprietary content. Adobe’s strength lies in its established position within creative workflows, giving it a potential advantage in embedding AI directly into tools that marketers and designers already use daily.

The integration of the team from Invoke, a generative media startup, into the Firefly Foundry group suggests Adobe is investing heavily in in-house expertise to stay competitive in AI-driven creative production.

For marketing professionals, the announcement signals a step toward a more controlled and brand-safe application of generative AI—one that sits inside existing creative ecosystems rather than outside them. While it remains to be seen how widely businesses adopt proprietary AI models, Adobe’s move underscores how content production is shifting from experimentation to operational deployment across the marketing industry.