Switzerland - AI Challenges Report

Overview

The Interdepartmental Working Group on Artificial Intelligence (IDWG AI), led by the State Secretariat for Education, Research and Innovation (SERI), submitted the report "Challenges of Artificial Intelligence" to the Federal Council on 13 December 2019. The document provides a cross-cutting assessment of AI-related opportunities and risks and maps 17 thematic areas and 37 priority action items for the Federal Administration. The report underlines Switzerland’s strong research base and innovative ecosystem, advocates maintaining a largely technology-neutral regulatory stance, and proposes targeted, sector-specific measures where existing rules require clarification. The report and its executive summary are published by SERI; see the SERI AI page and the official report files for full text and executive materials: SERI – Artificial Intelligence and the full report (German): Challenges of Artificial Intelligence (Bericht, 13.12.2019).

Definitions

The report adopts usage-consistent, policy-focused definitions rather than formal statutory definitions. It treats 'artificial intelligence' as a set of methods, especially machine learning and algorithmic systems, that enable automated data-driven inference, prediction, classification, decision-support or autonomous decisions. The emphasis is on socio-technical systems (algorithms plus data plus organisational processes) and on use-cases that have societal, economic or regulatory relevance. The report treats risk levels contextually—depending on application domain, societal impact and potential to affect fundamental rights—and recommends risk-proportionate responses that rely on existing legal principles such as non-discrimination, proportionality, and data protection.

Governance and Institutional Framework

The IDWG AI is explicitly interdepartmental: it brings together multiple federal offices to map policy needs and coordinate action. SERI led the work, with participation from departments responsible for economic policy, communications, transport, health, security, and justice. The report recommends establishing a national competence network and clearer coordination roles inside the federal administration, while delegating sectoral enforcement and operational responsibilities to existing specialised authorities. The Federal Council subsequently adopted "Guidelines on Artificial Intelligence for the Confederation" in November 2020—an implementation step that the report itself recommended—to provide orientation for federal offices and ensure horizontal policy coherence. Key institutional recommendations include appointing contact points within federal offices, strengthening the role of the Federal Office of Communications (OFCOM/Bakom) in cross-cutting review and coordination, and supporting competence hubs such as the proposed national network: SERI AI resources and public competence networks like the "Competence Network for Artificial Intelligence (CNAI)".

Key Focus Areas

The report clusters recommended actions in 17 thematic fields. High-priority areas include: (1) education and workforce development—strengthening AI curricula, reskilling and lifelong learning; (2) research and innovation—sustaining world-class research, computing and data infrastructures; (3) data policy and access—encouraging secure data-sharing infrastructures aligned with privacy law; (4) public administration—adopting guidelines for AI procurement, development and use within the Confederation; (5) sectoral regulation—assessing and adapting rules in health, mobility, finance and media to ensure safety, fairness and trust; (6) cybersecurity—addressing AI-specific threats and resilience; (7) ethics, transparency and fundamental rights—promoting traceability, explainability and human oversight; and (8) international engagement—aligning Swiss practice with EU and multilateral developments. The report stresses that Switzerland’s policy mix should preserve innovation-friendly conditions while ensuring safeguards for individuals and society. Further recommended activities include standardisation work, promotion of interoperability, and support for start-ups and SMEs to adopt AI responsibly.

Implementation Framework

The report recommends a pragmatic, multi-level implementation approach. At the federal level it calls for: (a) operationalising the 37 priority actions by allocating responsibility to specific offices; (b) developing shared services (secure data platforms, compute resources) that can be used across agencies; (c) issuing guidance for procurement and development of AI systems in public administration; (d) coordinating with cantons and the private sector; and (e) prioritising pilot projects and sandboxes to test regulatory approaches. The Federal Council’s November 2020 guidelines follow this model by offering an orientation framework rather than prescriptive rules. Implementation should follow risk-proportionate principles: light-touch measures where impact is low and targeted safeguards where systems affect safety, fundamental rights or critical infrastructure. The report also recommends that sector-specific regulators review existing laws and propose clarifications where AI-specific legal uncertainties impair enforcement or public protection.

Monitoring and Evaluation

To ensure adaptability, the report proposes continuous monitoring of AI developments and periodic evaluation of policy measures. Monitoring indicators include: research output and talent pipeline metrics, uptake in public administration, incidents affecting safety or rights, investments and start-up activity, and international regulatory developments. The report endorses creating fora and data collection mechanisms to aggregate these indicators and to report progress to the Federal Council. It also recommends staged evaluations for regulatory interventions and pilot programmes, using sandboxes and controlled trials to gather evidence before broader roll-out.

Penalties, Liability, and Appeals

The report itself does not create new penalty regimes; instead it recommends clarifying responsibilities under existing sector laws (e.g., health, transport, financial market regulation) and under Swiss civil and administrative liability frameworks. Where risks to safety or rights exist, the working group advises that enforcement should rely on competent sector authorities applying existing sanctions and corrective measures. The IDWG also flags potential gaps in liability allocation for autonomous systems and recommends further study and potential legislative clarification to ensure victims have access to redress and to assign responsibilities between developers, deployers and operators.

Relationship to Other Instruments

The report positions itself as complementary to the "Digital Switzerland Strategy" and other sectoral strategies. It recommends coordination with data-protection authorities and points to the applicability of the Federal Act on Data Protection. The IDWG explicitly calls for alignment with international initiatives and standardisation work and indicates that where international rules develop (notably EU-level regulation), the Federal Administration should assess implications for Swiss law and adapt where necessary. The report also spurred the Federal Council’s 2020 guidance for federal agencies and influenced subsequent sector initiatives and research funding priorities: see the SERI AI page and related SATW commissioned studies for thematic deep dives: SATW – Switzerland is in a good starting position.

International Alignment

Given Switzerland’s deep economic links to the EU and international organisations in Geneva, the report emphasises active engagement in multilateral fora and monitoring of EU regulatory developments (including the EU AI regulatory initiatives). It recommends preserving regulatory compatibility where feasible and participating in standardisation bodies. The IDWG advises the Federal Department of Foreign Affairs (FDFA) and relevant federal offices to strengthen international coordination, to protect Swiss interests in trade and research cooperation, and to contribute to the development of interoperable, rights-respecting international norms for AI. External resources and comparative analysis (for example, AI Watch summaries of national approaches) are cited to support alignment decisions: AI Watch – Switzerland.

Implementation Timeline

Phase	Actions	Indicative Dates
Initial acknowledgement	Federal Council acknowledges report; executive summary published	13 December 2019
Guidance and coordination	Federal Council issues orientation guidelines for Federal Administration (based on report)	November 2020
Operationalisation	Offices assign responsibilities; pilots and sandboxes; competence network formation	2020–2022 (ongoing)
Monitoring & evaluation	Regular reviews; progress reports; adaptation of actions	Annual/biannual reviews recommended

Sources and References

Source	Type
SERI – Artificial Intelligence (SBFI/SERI official page)	Primary Source
Challenges of Artificial Intelligence: Report (Bericht, 13.12.2019) – German PDF	Primary Source
Challenges of Artificial Intelligence: Executive Summary (English PDF)	Primary Source
European Commission / AI Watch – Switzerland AI Strategy Report	Secondary Source
SATW – Switzerland is in a good starting position (analysis)	Secondary Source