Promote an innovation-friendly approach to AI Governance || Combine a lean regulation safeguarding basic rights with sector-specific self-commitments
Switzerland pursues an innovation-friendly governance approach based on regulation designed to safeguard fundamental rights and ensure compliance with the Council of Europe’s AI Convention, thereby committing to internationally recognised standards of protection. This framework is supplemented by sector-specific voluntary commitments. Blueprints and guidance on the implementation of sectoral AI codes support the effective design of these voluntary commitments.
Actions
The actions may either build on and strengthen existing initiatives or constitute new
Implementation Toolkit for responsible AI self-commitments
Ensure effective self-committments for responsible AI as a key element of the Swiss Approach to AI Governance
Context (why)
Switzerland’s AI governance approach blends cross-sector legislation with non-binding measures to meet the Council of Europe’s AI Convention. One important element is self commitments.
To build public trust, the government provides a foundational "blueprint" for AI self-commitments. The Implementation Toolkit translates the blueprint into practical standards and tools to facilitate the development of industry-specific Codes of Practices.
Objective
Support the Swiss AI Governance approach by developing an Implementation Toolkit that complements the blueprint for self-commitments. This enables sectors to develop codes of conduct/practice for responsible AI still in 2026.
Key Elements
The Implementation Toolkit translates key principles into practicable standards, frameworks and tools. It is structured along the AI life cycle stages addressing different organisational actors across the stages of:
Design → Data → Model Development → Validation → Deployment → Monitoring
Contributors
Coordination between OFCOM, digitalswitzerland and ETH Zürich will take place to strengthen the practicability of the blueprint through an Implementation Roadmap and Toolkit developed by digitalswitzerland and ETH Zürich
Target Group of the Action
Sectoral associations
National AI Innovation and Regulatory test environments (Sandboxes)
Create terms of establishment and use for national AI Innovation and Regulatory sandboxes and define the legal requirements.
(Cross-Topic Action: Research & Innovation | Public AI Governance)
Context (why)
Switzerland faces a critical gap in both technical and regulatory testing infrastructure, leaving AI stuck in "labs" and unprepared for real-world application. Without testing environments, startups lack the resources to scale, while established industries remain paralyzed by liability risks and the complexity of emerging digital laws.
Objective
Create terms of establishment and use for national AI regulatory and innovation sandboxes and define the legal requirements until end of 2026
Key Elements
Combine Technological Sandboxes (physical infrastructure/digital twins) and Regulatory Sandboxes (with authority oversight) under a single umbrella, while maintaining distinct organizational structures and participation terms.
Integrate sandbox approaches into hard law to provide clarity on regulatory treatment.
Enable testing with temporary exemptions to prevent innovation-stifling legislation for selected regulations in the digital field.
Establish a dual-track sandbox architecture by integrating horizontal (cross-sector) and sector-specific testing environments to ensure clear governance and jurisdictional alignment at the national level.
Contributors
Target Group of the Action
AI product developers
Constitutional Digitalisation Article
By introducing a Digitalization Article into the constitution, a new federal competence is to be established, including specific rules for the participation of cantons, cities, and municipalities. This will ensure that, in the future, essential standards such as those for data transfer between authorities are binding for all public entities.
Context (why)
Current digitisation and AI barriers include a lack of technical, organizational, and legal standards required for interoperability across all state levels. The existing competence order is no longer sufficient for the digital age.
Objective
To establish a federal competence that allows for binding national AI/digital standards and solutions while ensuring federalist participation.
Key Elements
1. Legislative Draft: Following Motion 26.3005 to draft a flexible constitutional article.
2. Interoperability: Defining standards that allow for "multi-use" of solutions across federal levels.
Contributors
Target Group of the Action
AI product developers
Other topics
Other Topics of the AI Action Plan for Switzerland
Scaled AI Education and Literacy
Creating an AI Competency Boost for our economy and the entire population.
World-Class Research and Innovation
Expanding our world-class research and innovation through close European cooperation.
Resilient Digital Infrastructure
Building a resilient and competitive digital infrastructure for Switzerland as a business location.
AI-Ready Data
Unlocking high-quality AI-ready data as fuel for research, innovation and business models.