Skip to main content

AI trends seminar 2026: From vision to reality – shaping the next wave of AI

Monday, February 23, 2026

The AI landscape is moving beyond the experimental stage. During this year’s AI Trend Seminar 2026, experts from organizations including Google Cloud, Recorded Future, RISE, and Zenseact gathered to discuss the path toward digital sovereignty and reliable autonomous agents.

The seminar focused on three primary trends: the de-globalization of AI, breakthroughs in physical intelligence, and the journey toward reliable agentic workflows. Led by moderators Astrid Sjögren, Strategic Initiative Manager for SMEs at AI Sweden and Daniel Gillblad, Chief of AI at Recorded Future and Senior Advisor at AI Sweden, the panel discussed how Sweden and Europe can secure their competitiveness in an increasingly complex world.

Trends Selection

This year’s selection of trends marks a clear shift from speculative future visions to areas where AI is now creating actual value in the real world. Mauricio Muñoz, AI Research at AI Sweden highlighted how the technology has matured, particularly in the natural sciences, where models are now demonstrating an impressive ability to solve complex mathematical problems. 

Another central insight involved the ongoing commercial transformation. As the costs of advanced models drop rapidly, high-level intelligence is becoming a commodity. This fundamentally changes the playing field for companies, making it possible to scale the use of powerful AI in ways that were previously economically unsustainable.

Watch this segment in the full webinar → 

De-globalization of AI and Geopolitics

The panel, featuring Irene Ek, Public Policy Manager at Google Cloud, Malin Frenning, CEO of RISE, and Ödgärd Andersson, CEO of Zenseact, analyzed how global tensions and decreased reliance on the U.S. are increasing the demand for European sovereignty. The discussion focused on the need for European digital sovereignty and what is required for Sweden to remain secure and competitive.

The experts emphasized that global ecosystems are becoming increasingly regionalized. We are currently seeing how data regulations are forcing the development of locally adapted products, particularly, within the automotive industry. To counter strategic vulnerabilities, the panel stressed that it is critical for Europe to move from being a mere consumer of technology to becoming an active developer by investing in its own data centers and independent infrastructure.

Watch this segment in the full webinar →

Embodied Intelligence and Physical AI

Amy Loutfi, Professor at Örebro and Linköping Universities and Program Director at WASP explained why physical AI - known as Embodied AI - is this year’s megatrend. The focus has shifted from simple physical robots to integrated machine learning, where perception and action merge in complex environments.

By using new models capable of interpreting three-dimensional environments, machines are learning to understand the laws of physics and how different actions affect the world around them. To ensure the technology is safe, these systems are often trained in advanced digital simulations. In these environments, they can practice handling rare and dangerous events, such as traffic accidents, in a controlled virtual environment before being deployed on our streets or in our factories.

Watch this segment in the full webinar → 

Agentic Systems: The Path to Real-World Success

Aron Lagerber, Principal AI Scientist at Recorded Future discussed the path toward autonomous systems with over 99% reliability in large-scale enterprise environments.

His conclusion was that reliability is the most critical feature. Achieving this requires a new approach to agent design. By introducing strict traceability and separating the model from the system, developers can build agentic systems that adapt to—and grow with—the technology of tomorrow.

Watch this segment in the full webinar ->

You might also be interested in

Mauricio Munoz portrait picture

Column: Insights from the world’s premier AI conference

2026-01-23
In early December 2025, top researchers in artificial intelligence (AI) and machine learning (ML) gathered for the Conference on Neural Information Processing Systems (NeurIPS). The conference is one...
Mikael Ljungblom, Rosenbad i Stockholm med svenska flaggan

AI Sweden on the Government’s National AI Strategy: "High ambitions demand bold leadership and substantial investment"

2026-02-20
The Swedish government today presented a national AI strategy featuring ambitious goals for the country. Mikael Ljungblom, Director of Public Policy and International Relations at AI Sweden, welcomes...
Woman (Moa Tivell, AI Sweden) next to AI Sweden report cover "Framtidens arbete" featuring yellow geometric cubes.

New report on AI's impact on the labor market: skills development more vital than urgent transition

2026-02-17
Artificial intelligence’s structural impact on the Swedish labor market is primarily characterized by evolving job tasks and new skills requirements—rather than the disappearance of jobs. This is...