Adobe Inc. has launched a new enterprise offering called AI Foundry, which enables businesses to co‐develop custom generative AI models trained on their own branding assets and intellectual property. These bespoke models are built on the company’s existing Adobe Firefly family of generative models, which were trained entirely on licensed data and operate across text, image, video and 3D‐scene generation. The key benefit: enterprises get models that reflect their unique style, tone and IP while leveraging Adobe’s AI infrastructure—Adobe retains ownership of the base models but fine-tunes them for each client. Pricing is usage-based rather than per-seat, and early adopters include major brands. The announcement signals Adobe’s deeper push into the enterprise AI space, positioning itself as a trusted vendor offering both creative workflow automation and brand-safe generative AI.
Sources: TechCrunch, AI Economy
Key Takeaways
– Enterprises now have the option to deploy generative AI models tailored to their identity, trained on their own assets (not generic models), reducing reliance on off-the-shelf AI.
– Because the models are built on Firefly (trained on licensed data) and fine-tuned specifically per client, Adobe is selling the service as “brand‐safe” and IP-aware.
– The pricing model shifts away from traditional user seats toward usage-based billing, aligning cost with output rather than headcount.
In-Depth
In a move that underscores the growing importance of generative AI in enterprise workflows, Adobe has rolled out its new AI Foundry service, which offers businesses the ability to co-create and deploy generative AI models that are intimately aligned with their brand, intellectual property and creative guidelines. Built atop Adobe’s Firefly model family, this service reflects a strategic shift: rather than offering generic “one-size-fits-all” AI models, Adobe is positioning itself as the AI partner for enterprises seeking differentiated capabilities and the assurance of brand coherence. The Firefly models, introduced earlier, were themselves notable for being trained solely on licensed data—addressing many of the concerns around copyright and model provenance. Now, with AI Foundry, Adobe takes the next step by fine-tuning those models with a company’s proprietary media, assets, voice and style — meaning the generated output is distinctive and aligned to the enterprise’s identity.
From a business perspective, several implications stand out. First: operational efficiency. Creative teams often spend large amounts of time creating variants of marketing assets, adapting them for different channels, formats or languages. A brand-specific generative AI model could significantly accelerate that process, enabling faster turnarounds and higher volume without sacrificing brand fidelity. Second: risk management. Because Adobe emphasizes the licensed base data and separation of the client-trained model from its base, enterprises concerned about copyright, model safety and output consistency may find value in a vendor like Adobe rather than relying on open-source or generalist AI providers. Third: economic model innovation. Usage-based pricing (rather than per-seat licensing) may better reflect the value—brands pay for output (assets generated) rather than just users logging in. That aligns cost to volume of creative work.
Of course, there are open questions. How cost-effective will this service be for smaller firms? For large enterprises it may make sense, but mid-market companies might still face high bar of adoption. Also, even if the base model is safe and licensed, the value of training a model on proprietary data depends on the volume and quality of that data; not every enterprise will have sufficiently rich asset libraries. And generative AI still faces broader issues: hallucinations, brand-consistency drift, and the need for human oversight remain. Adobe acknowledges this by positioning AI Foundry more as a partnership with applied scientists and creative workflow experts rather than a plug-and-play product.
For Adobe, this service helps differentiate it in an increasingly crowded generative AI market. As other providers focus broadly, Adobe’s enterprise pitch is about creative workflows, brand identity, and deep integration into its existing creative stack (Creative Cloud, Experience Cloud, etc.). This could reinforce its position with agencies and brands already using its tools. In sum, AI Foundry represents both an evolution of Adobe’s AI strategy and a recognition that enterprises want generative AI that is tuned, safe, brand-aware and integrated. While generative AI remains transformative in promise, the next phase is less about broad access and more about customisation, control and alignment with business identity—and Adobe appears to be angling to lead in that direction.