AI-Generated Images: Open Wonder Helps Established Companies Maintain Consistent Brand Communication

Using AI optimally remains a major challenge for many mid-sized companies as well as large corporations. This is particularly noticeable to the general public when it comes to visual language. Individual AI-generated images often look impressive, but across markets, channels, and formats, inconsistencies quickly emerge. This is where Open Wonder comes in: the team translates its clients’ brand knowledge into a machine-readable “Brand Layer,” ensuring that every AI output consistently reflects the brand. Tim Herzog, Co-Founder and CEO of Open Wonder, gives an exclusive interview and shares insights about the current collaboration with PSI Software SE – a leading provider of software for controlling critical infrastructure.

Creating images with AI today seems incredibly simple and yet simultaneously challenging. Anyone can participate, yet the results are often unsatisfactory. Why do major partners choose to work with you?

Because generic AI tools don’t understand a brand. FLORA or Google’s Nano Banana 2 deliver impressive individual images, but when an international brand needs hundreds of them consistently across all channels, markets, and formats, inconsistencies scale. Open Wonder translates a company’s brand knowledge into a machine-readable structure – our Brand Layer. Every output is generated within the defined brand logic. For our clients, leading brands in e-commerce, healthcare, and consumer goods, this means scaling internally without losing control over their own brand.

Evoto (c) Manuel Gutjahr

How and when was your company founded, and how do you connect with decision-makers at such renowned companies?

Open Wonder was developed by our team at AN®. Since 2018, we have been doing brand work for international companies – strategy, identity, digital platforms – and we always saw the same pattern: brands are strategically well-defined but fail in operational execution. Generative AI has intensified this challenge. We then automated our own execution layer, which led to the creation of our product. Our first contacts came from our existing network. We speak the language of brand and marketing decision-makers, not just IT, which opens doors even at large corporations.

What differentiates your product from standard AIs like ChatGPT?

ChatGPT or Claude are powerful language models, but they are not designed for image creation. Even state-of-the-art image generators like Google’s Nano Banana 2 deliver impressive results, but they have no concept of brand. Open Wonder is designed with the brand in mind. Our Brand Layer captures a brand across eight dimensions, including color, typography, and tone, and makes this knowledge usable for generative systems. We are foundation-model-agnostic, using the best available model for each use case.

But the real difference lies elsewhere: Open Wonder creates a central place where all brand knowledge is operationally available. The brand team defines once how the brand should look and communicate. After that, anyone in the company can work with it – even salespeople, product managers, or local subsidiaries who are not designers. The output remains on-brand without anyone having to write a briefing or wait for the creative department.

Additionally, we integrate into existing enterprise processes: GDPR-compliant, EU infrastructure in Frankfurt, SSO, connection to DAM and CMS systems.

Can you tell us more about a collaboration you find particularly exciting? What were the challenges, and how did you overcome them together? Did you notice any “cultural” differences?

An exciting collaboration is with PSI Software SE, a leading provider of software for controlling critical infrastructure. The requirements were high-technologically, strategically, and in terms of brand consistency. PSI faced the challenge of creating an entirely new visual world and making it scalable and internationally deployable. An inconsistent appearance could quickly undermine credibility. The biggest cultural challenge: speed. We think in weeks, a publicly traded tech company in longer cycles. You have to learn that. Our approach was intentionally strategic: positioning, structure, and a modular design system as a foundation. On that basis, we systematically developed the visual world with Open Wonder and set it up as a scalable system. Today, image production there is consistent and used across various applications.

What are common points companies overlook when generating or using AI images? What risks come with it? Are there prejudices against your product? Where is awareness still needed?

Companies underestimate how quickly generic AI can dilute their brand. If every department and market independently uses FLORA or Nano Banana, visual noise emerges. Individually, images look good, but together they do not tell a coherent story. Legal risks also arise: copyright, data protection, and where data is processed. A common prejudice: AI images are a toy. But Open Wonder does not replace creatives—it provides them with infrastructure. The brand team defines the guardrails, and the system works within them.

How do you stay ahead of standard AIs and maintain your USP?

We consciously do not compete with foundation models. We build the intelligence layer on top. Our Brand Layer sits between the models and the output. If a better model appears tomorrow, we integrate it – the brand understanding remains. The way we formalize brands reflects 15 years of work and hundreds of projects. This is not easily replicable. In the long term, we are moving toward an agentic Brand Operating Platform that operationally manages brands.

What current trends or challenges do you see in your industry?

Foundation models improve weekly. This puts companies under pressure. Many know they need to act but don’t know how. What is missing is structure and a partner to discuss challenges with. We understand the issues and know how to solve them. Then there is the AI Act: for us as an EU-first company, an advantage. And the most important trend: the shift from tools to agents. The agent of the future understands the entire brand context, plans campaigns, and delivers assets directly into the systems teams work in. That is exactly what we are building.

Contact

Website: www.openwonder.com

LinkedIn: Tim Herzog