Agents & Agency –
The New Operating Model

Europe's AI sovereignty needs an economic foundation. Our task is to build technology that can compete effectively. Europe's task is then to deploy it on a large scale, not just pilot projects. Only in this way can we shape the technology we will depend on tomorrow.

AI is evolving into an autonomous actor. Governance, control, and transparency are therefore becoming a crucial competitive advantage, as this is the only way to achieve secure and scalable value creation.

Being successful with AI has almost nothing to do with AI itself. It has to do with us humans.

AI models are becoming interchangeable. The real competitive advantage comes from company knowledge. We are building the system in which this knowledge remains protected and becomes productive – sovereign from Europe, for organizations that want to use AI securely and effectively.

An AI model is quick to build – the real work lies in providing it with the right data at the right time during operation.

If you don't control your agents, your agents control your company.

A PDF is not governance. AI agents have been acting for a long time. Do you know which ones are making decisions in your company?

We talk a lot about automation when we imagine AI, but I think its bigger impact may be giving everyone their own Chief of Staff, available for cents a month, connecting information, spotting problems early, helping people make better decisions. That is already starting to happen. The business value is not, not yet in individual output, customer experience, or revenue. Mostly, still, it shows up as a cost

Models are becoming mass-produced, but implementation is not. Until 2027, the advantage lies in the data foundations, redesigned workflows, and organizations that can actually integrate the technology on a large scale.

How a company uses AI systems in 2027 will determine its competitiveness. Those who maintain control over the context and the decision-making logic themselves will remain capable of acting, even if model providers change or fail.

Physical AI only scales with real added value if AI agents not only learn movements but also understand machines and their technical interrelationships. This requires transforming scattered engineering and product data into reliable, usable product understanding.

Physical AI doesn't learn in the lab. It learns in real-world operations. Every part is different, there's no fixed sequence, no script. The more stations are running in actual production, the more data is generated, and the faster the path to full autonomy.

The year 2027 will be the decisive turning point for companies that manage the transition to successfully and profitably deploy AI in their core value-creating processes.

The next decade of AI will not just be won by those who built the underlying technology, but by those who built the workforce capability, and the institutions, to benefit from it.

Europe talks about AI sovereignty, while its most successful AI startups run on US models and US infrastructure. As long as we don't bridge this gap between ambition and market reality, sovereignty will remain just a buzzword.

Sovereignty and resilience determine whether companies merely use AI or strategically master it. Technological capability thus becomes a key prerequisite for AI value creation.

Guided by "Be Digital. Stay Human.", I see technology as a means to empower people and strengthen trust in a digital world. Innovation creates real value when it delivers benefits and is based on strong values.

Every week that a workforce doesn't use AI is a week of lost experience. Competitiveness doesn't depend on who buys the best technology, but on who learns fastest as an organization.

The future of retail is not about helping customers find products but helping them find inspiration. AI becomes truly valuable when it transforms transactions into engaging, personalized experiences.

Future-proof AI systems are modular and rely on open standards: Those who can exchange models and components at any time remain sovereign – both technologically and economically.

The race for AI is in full swing, and we in Germany still have a chance. However, only if we finally transform ourselves from tool tourists into application world champions!

Pilots impress. Harmonized, scaled processes deliver EBIT.

The AI Labs' AI race is not our race. Our race is the most meaningful adoption, with measurable value.

We not only need to learn how to deal with AI, but also develop our own decision-making style! Because: Those who decide shape the future. Those who hesitate react to it.

Following the widespread adoption of Generative and Agentic AI, companies now face the challenge of how and in which areas real AI value creation can take place.

Most organizations are not losing to AI. They are losing their own ability to decide.

The question is no longer whether companies will use AI, but whom they will trust. Those who don't control access to, rules for, and pricing of their models have already relinquished control.

The AI transformation offers the BMW Group the opportunity to establish a new operating model and to efficiently scale AI agents – as a basis for sustainable value creation.

AI is not merely an efficiency tool, but a driver of change. It delivers the greatest added value when companies rethink their business and customer experiences – through the interplay of people, data, and AI.

AI is the topic of the moment, a great opportunity to drive innovation on a massive scale. Let's make Germany an AI nation!

Those who want to make good decisions for themselves and their company need to understand more than just the current state of technology. At the AI Summit, we'll discuss the developments that are now possible – and how we can seize these new opportunities.

AI leadership means taking the lead. Making your own path visible. And demanding that everyone in the company rethink work with and through AI. Those who delegate AI lose. Those who lead it win.

The business model of the past is no longer viable – now we must adapt processes and organizations to the age of agents in order to secure Europe's leading economic position.

The next phase of the AI revolution is beginning, with more power, more opportunities, but also higher risks.

Artificial intelligence is developing rapidly. It is crucial to learn from each other and share experiences early on.