March 19, 2026
L’Oreal is now producing up to 50,000 images and more than 500 videos every month using internal generative AI tools built with Google Cloud. The beauty giant’s AI content lab, dubbed “Creaitech,” uses Google’s Imagen 3 and Gemini multimodal models to streamline marketing production across its portfolio of 37 brands.
The result: a 30% increase in content production capacity while maintaining brand quality standards – and a blueprint that any marketing team can learn from.
Inside L’Oreal’s AI Content Factory
The Creaitech platform works as a creative accelerator. Marketing teams describe an idea using prompts, and the AI generates visuals including product shots, packaging redesigns, storyboards, and concept art. Google’s Veo 2 video model can then turn static images into eight-second animated sequences – useful for social media content that would traditionally require video shoots.
L’Oreal’s teams use the platform for several specific workflows:
- Product visualization: Testing pack shots in various locations and settings without physical photography
- Concept development: Generating creative concepts and mood boards for campaign ideation
- Packaging design: Rapidly iterating on packaging redesigns before committing to production
- Social content: Adapting visual assets and video footage for different platforms and regional markets
The platform also integrates with Adobe Firefly for additional creative capabilities, giving teams multiple AI tools within a single production workflow.
The Line They Will Not Cross
What makes L’Oreal’s approach notable is where they draw the line. The company has explicitly decided not to use generative AI to create images of people for marketing campaigns or external communications. This includes “life-like” face, body, hair, and skin images for product advertising.
For a beauty company that sells products based on how they look on real people, this is a significant ethical stance. AI handles the creative heavy lifting – product shots, environments, packaging, concepts – but human models and real photography remain central to consumer-facing campaigns.
L’Oreal has also built structured governance workflows including environmental impact assessments, data usage limits, and internal oversight for AI experimentation. This kind of framework is increasingly important as brands face scrutiny over AI-generated content authenticity.
Key Takeaways
- L’Oreal produces 50,000 AI images and 500+ videos monthly via Creaitech platform
- Built on Google Imagen 3, Gemini, and Veo 2 video model
- 30% increase in content production capacity across 37 brands
- Will NOT use AI to generate images of people in marketing campaigns
- Structured AI governance with environmental impact and data usage oversight
What This Means for Marketers
AI content production is now enterprise-grade. L’Oreal is not experimenting – they are running 50,000 AI images per month in production. If the world’s largest beauty company trusts AI for visual content at this scale, the technology is ready for most marketing teams.
Draw your own ethical lines early. L’Oreal’s decision to ban AI-generated people images sets a precedent. Define what AI can and cannot do in your brand’s content strategy before you scale up – not after a controversy forces your hand.
Product photography is the first to be disrupted. Pack shots, environment staging, and packaging mockups are the low-hanging fruit. If your team is still booking physical photo shoots for product-only content, you are likely overspending compared to AI alternatives like Firefly, Imagen, or Midjourney.
Sources: Retail Dive, Google Cloud Blog, Cosmetics Business, AI News. This article may contain affiliate links.
For a deeper look at the tools shaping this space, see our Best AI social media tools 2026 guide.