Colombus Consulting, in partnership with Oracle and the Geneva School of Business Administration, has published the 2026 Observatory of Data and AI in Switzerland for the third consecutive year.
The study reveals that AI is now firmly established in Swiss companies’ strategies: it is entering long-term plans, internal teams are taking shape, and a new wave is emerging with Agentic AI. However, scaling up remains held back by data quality, governance and organisational transformation. Swiss companies are no longer asking whether AI is strategic, but how to integrate it sustainably into their operating model.
This observatory has the same objectives as in previous years: to provide decision-makers with a navigation tool to understand the present, anticipate future changes and compare their progress with that of their sector or the market.
The panel surveyed covers the whole of Switzerland and includes companies from all sectors of activity. The Observatory uses the same methodology as in previous years, organised in several stages:
- The survey, via a detailed questionnaire, which was answered by more than a hundred organisations, covering their governance, technological maturity, and their ambitions and obstacles in their data transformation projects;
- The additional collection of qualitative insights via in-depth interviews with experts and opinion leaders;
- On this basis, the formulation of observations and lessons learned, the main ones of which are listed below.
Key results:
Strategic conviction is now established, but execution still lags behind
- 81% of organisations believe that AI can help solve the complex problems facing their industry
- 47% (+12 pts) integrate AI into their long-term planning, but only 9% truly place it at the core of their strategy — a ‘knowing-doing gap’ that no longer reflects a lack of conviction, but a lack of structuring
Data foundations are progressing, but unevenly
- 82% of organisations remain at a low or intermediate level of Data & AI maturity, and 51% (-10 pts) rate their data quality as good to excellent — a decline that reflects rising requirements rather than an actual deterioration, as use cases grow more ambitious
Leadership plateaus after the 2025 leap, while internal capabilities keep strengthening
- 55% (-8 pts) of leadership teams report a good to high level of AI knowledge, while 65% (+8 pts) of organisations now have at least a small in-house AI team
Organisational transformation remains one of the main obstacles to the full integration of AI
- 82% consider Change Management critical to the success of AI, but only 11% are actively redesigning roles around human-AI collaboration
Generative AI is refocusing, while Agentic AI enters an active exploration phase
- 49% (-7 pts) have identified use cases and are running active GenAI pilots, and 46% (-29 pts) have deployed assistants at scale — a decline that reflects a refocus on value
- 69% are already exploring or experimenting with Agentic AI, but 90% keep autonomy at a ‘zero’ or ‘human-in-the-loop’ level
Ethics and sovereignty are becoming operational conditions
- 72% (+4 pts) incorporate ethical considerations into their AI-related decisions, and 62% actively monitor their level of sovereignty or consider it a blocking factor
However, training remains a point of vigilance
- 34% (+14 pts) still provide no formal training on these issues — even as AI spreads well beyond expert teams
Industrialisation remains selective
- Only 16% report having deployed AI at scale, and 36% still do not measure any tangible value from their AI projects
