Geopatriation, cybersecurity & physical AI: AI in the year 2026

Last updated: 04.12.2025 09:00

Everyone talks about AI and thinks of ready-to-serve chatbots and voicebots on websites. But artificial intelligence encompasses much more and creates possibilities that many people haven't even thought of yet. Here is a brief outlook from VIER Vice President Carsten Fiegler on what we can expect as early as 2026.

In 2026, we will experience the next evolutionary stage of intelligent automation: distributed AI agents that work together autonomously like a team and whose work only occasionally needs to be corrected by a human - known as Human in the Loop. These multi-agent systems (MAS) should enable companies to make complex business processes more scalable and resilient. Domain-specific language models (DSLMs), i.e. industry or function-specific language models, will also become important from now on. They are increasingly replacing generic AI systems, delivering more precise results, reducing incorrect outputs and enabling regulatory-clean prompting in sensitive environments.

Data: Where from and where to?

Given the tangible geopolitical tensions, many companies are already moving data and workloads away from global hyperscalers to sovereign or regional clouds. This so-called geopatriation will increase significantly in 2026. As a result, data sovereignty will become a strategic location factor. At the same time, digital provenance, i.e. the origin and integrity of data, models and content, will become a new competitive factor. This is because companies need mechanisms for tracking and authenticating digital assets in order to ensure trust in AI systems.

Security for and through AI

This is where new AI security platforms come into play. These new security platforms are designed to secure the entire life cycle of AI applications - from model training and data quality to policy management and attack detection. Security is thus becoming an integral part of the AI stack. This also includes the development towards so-called pre-emptive cybersecurity. This means that security strategies are no longer just designed to react to attacks, but increasingly include predictive detection and deception tactics. In future, AI-supported analyses will neutralize threats before they cause any damage.

Let's get physical!

One thing is certain: in 2026, artificial intelligence will leave the digital realm. In future, physical AI will control robots, drones and machines in the real world. IT, OT and IoT will merge into a common intelligence layer.

And: software development is becoming AI-native. Platforms will integrate generative and assistive AI functions directly into the development process. AI-native development platforms will significantly improve productivity, quality and time-to-market for companies!