Adobe Launches Firefly Foundry for Custom Enterprise AI Models

Adobe (Nasdaq: ADBE) launched Adobe Firefly Foundry, enabling enterprises to deploy proprietary generative AI models trained on their intellectual property across image, video, audio, vector, and 3D content.

Built on Adobe Firefly, Firefly Foundry enables custom enterprise models for multimodal content creation as content demand continues to grow rapidly. Adobe’s AI specialists assist with model training and integration, with early adopters including Walt Disney Imagineering, which is using the technology to develop brand-specific AI experiences.

Adobe Firefly Foundry allows organizations to train AI models using their existing catalog of brand assets. This helps ensure consistent output across marketing campaigns, product lines, and customer experiences.

The platform supports multiple content formats, allowing teams to generate materials while maintaining brand identity and accelerating production timelines.

Companies using the company’s creative technology platform have reported notable efficiency gains. The firm revealed it reduced its Black Friday campaign time-to-market by 60 percent through automated workflows.

Global brands such as The Coca-Cola Company, The Estée Lauder Companies, PepsiCo, and the National Football League are also leveraging these capabilities to accelerate campaign development and creative production.

Media and Entertainment

Media studios can use franchise assets to generate consistent material across formats, including characters, environments, and visual effects while maintaining story continuity. Creative teams are already exploring these capabilities to ensure consistency across projects.

Retail and Consumer Goods

Retail and consumer brands can scale localized campaigns, product visuals, and seasonal materials more efficiently. One global beverage company has partnered with a creative technology provider to build a design intelligence system that embeds brand rules directly into creative workflows.

Performance Marketing

Marketing teams can generate multiple ad variations for A/B testing and audience targeting. The platform integrates with advertising platforms such as Amazon Ads, Google Marketing Platform, Innovid, LinkedIn, and TikTok to support automated campaign activation and optimization.

Product Development

Manufacturing and consumer electronics companies can use the platform to create product renderings, packaging designs, and technical documentation using systems trained on proprietary design frameworks and engineering specifications.

Proprietary Development

Organizations can use their own intellectual property to train custom systems. These systems capture brand-specific attributes such as visual styles, character relationships, environmental details, and design system rules, enabling accurate outputs with minimal post-production adjustments.

Centralized Orchestration

The platform includes a unified system for development, testing, and deployment. Organizations can manage permissions, track usage patterns, and ensure results align with brand standards. It also integrates with asset management and collaboration tools to streamline workflows between creative teams.

Expert Collaboration

Technology specialists and data engineers work alongside enterprise teams to identify high-impact use cases and design implementation strategies. This collaborative approach helps organizations deploy advanced automation solutions faster and improve creative productivity.

Ecosystem Integration

The initiative integrates across enterprise innovative tools and marketing workflow applications. It introduces a beta AI-powered Production Agent that converts marketing briefs into channel-ready materials aligned with brand guidelines.

A suite of APIs also enables automated integrations such as video reframing, object compositing, and authenticity verification for scalable and reliable output.

Commercial Safety and IP Protection

The platform emphasizes transparency and intellectual property protection, allowing organizations to retain full ownership of their training data and generated outputs.

All generated material includes provenance signals and guardrails to support responsible technology use and regulatory compliance. The models are trained on licensed material, stock assets, and public domain data, helping reduce potential copyright risks.

Market Adoption and Availability

The company reports that 99 percent of Fortune 100 companies use advanced automation features in its applications, while many enterprise clients are adopting data-driven technologies to manage growing production demands.

The initiative is currently available to select enterprise customers. Hannah Elskar said it gives organizations access to advanced infrastructure, research, and expertise to build custom models.