Resilient Digital Infrastructure

Manage dependencies and strengthen capacity of AI systems || Ensure basic critical infrastructure for data processing and storage

Switzerland is securing its digital sovereignty and operational resilience by proactively managing digital dependencies and strengthening the capacity of AI systems. To sustain its position as a global innovation leader, Switzerland is prioritizing the establishment of robust, critical infrastructure for data processing and storage, ensuring that the foundations of the digital economy remain stable and scalable.

Actions

The actions may either build on and strengthen existing initiatives or constitute new

Define procurement standards to manage public digital infrastructure dependencies

Develop procurement standards for resilient AI infrastructure across different state levels.
Context (why)

The federal administration has already established a strategic foundation through the mandatory "Schutzbedarfsanalyse" and the DTI’s cloud classification frameworks. Additionally, the promotion of Open Source architectures can help reducing vendor lock-in. Nevertheless, procurement offices still lack harmonized operational metrics and enforcement aids (Vollzugshilfen) to navigate the selection of resilient services for critical workloads. This creates a fragmented landscape where authorities beyond the federal level must balance the risks of cloud and AI adoption without a unified methodology for choosing the right degree of resilience.

Objective

In 2027 introduce a legally embedded framework for critical infrastructure and critical public services within procurement processes that defines multiple dimensions of digital resilience requirements.

The framework enables authorities to mandate high-resilience hosting depending on the criticality of data. This includes requirements to demonstrate how dependencies are managed.

Key Elements

Including different dimensions of resilience:

  • Strategic resilience

  • Legal resilience

  • Operational resilience

  • Cyberresilience

  • Data Encryption

Contributors

Target Group of the Action

Procurers of critical infrastructure  and critical public service

Increase AI energy efficiency / transparency

Develop a full-stack labeling and disclosure framework for AI infrastructure, models, and services, building on the SDEA (Swiss Datacenter Efficiency Association) Label.
Context (why)

For an energy-importing nation, maximised energy efficiency of AI datacenters is crucial. The ability to do maximise Compute-Throughput isn't just an environmental goal but rather a prerequisite for technological sovereignty.

Objective

Developing a full-stack labeling and disclosure framework covering infrastructure, models, and services.

Key Elements

Switzerland should build on the pioneered SDEA (Swiss Datacenter Efficiency Association) Label, which is supported by the Federal Office of Energy.

Contributors

Target Group of the Action

Data center operators

AI model providers

Forecast AI compute and electricity demand in Switzerland

Identify and quantify compute and electricity requirements for critical workloads to secure Swiss data center capacity for national digital resilience.
Context (why)

To safeguard national digital resilience, Switzerland must secure sufficient domestic compute capacity for its critical workloads. High-density compute is becoming a "strategic constraint". Data centers will soon hit the ceiling of what local power grids can provide. Identifying and then securing the compute and electricity capacity is thus crucial.

Objective

By mid 2027, identify and quantify the specific compute and electricity requirements of Switzerland's critical workloads to enable securing sufficient Swiss data center capacity for national digital resilience.

Key Elements
  1. Critical workload compute identification

  2. Assessing current Swiss data centers to determine how much high-density space (GPUs/HPC) is available

  3. Identify the compute gap and translate into the electricity gap needed to cover demand.

Contributors

Target Group of the Action

Data center operators; Energy grid operators

Ideal Framework conditions for data center development

Establish a regulatory and spatial planning framework to identify priority zones and fast-track efficient data center development.
Context (why)

As the backbone of the AI economy, the demand for local data center capacity is skyrocketing. The development of data center projects can lead to conflicts over land use, grid instability, and missed opportunities for heat recovery. Establishing proactive conditions for spatial planning ensures that infrastructure is built where it is most efficient and least disruptive, providing investment security for developers while protecting community interests.

Objective

Establish a comprehensive regulatory and spatial planning framework for the national and cantonal level that identifies "priority zones" for data centers, mandates a minimum energy reuse factor and streamlines the permitting process to reduce lead times.

Key Elements
  1. Mapping designated areas with existing high-load power access and proximity to district heating networks to minimize infrastructure build-out.

  2. Projects located in these areas undergo "fast-track" mechanisms including Transmission System Operators (TSOs) to ensure grid stability.

  3. Facilitate respective energy grid extension and energy allocation in these areas.

Contributors

Target Group of the Action

Data center operators, Permitting authorities

Introduction of the e-ID and further development of the trust infrastructure

Establish the Swiss e-ID and develop the trust infrastructure to enable public and private digital services.
Context (why)

The e-ID is a key component of Switzerland’s digital transformation. The e-ID is issued by the federal government and can be used by private individuals, authorities and companies.

The further development of the e-ID is to address connection to international trust networks, backup services and the issuance of the e-ID in third-party wallets.

Digital Identities in general will be fundamental to address emerging challenges around AI-driven interactions. As AI agents increasingly act on behalf of individuals and organisations, the question of who controls, and is accountable for, those actions becomes critical.

Objective

The aim is to enable Swiss residents and Swiss nationals living abroad to identify themselves securely digitally.

The adoption should be taking place broadly and with a steep uptake and adoption.

The underlying trust infrastructure that enables cryptographic proof of identity and controllership.

Key Elements

State-issued e-ID: digital identity stored in a wallet on the holder's device

Trust infrastructure: decentralised registries and protocols for verifying issuers, holders, and verifiers.

Verifiable credentials: cryptographically signed attestations (diplomas, permits, mandates) presentable without contacting the issuer.

Privacy by design: selective disclosure, data minimisation, zero-knowledge proofs.

Legal framework: BGEID and ordinances, aligned with EU Digital Identity Wallet architecture.

Contributors

Target Group of the Action

Swiss residents, Swiss nationals abroad

Federal, cantonal & communal authorities

Private-sector service

Providers

AI system operators & deployers

Resilient Digital Infrastructure Topic Lead:

Yves Zischek

Yves Zischek

Managing Director, Digital Realty

Other topics

Other Topics of the AI Action Plan for Switzerland

Scaled AI Education and Literacy

Scaled AI Education and Literacy

Creating an AI Competency Boost for our economy and the entire population.

Education and Literacy Actions
World-Class Research and Innovation

World-Class Research and Innovation

Expanding our world-class research and innovation through close European cooperation.

Research and Innovation Actions
AI-Ready Data

AI-Ready Data

Unlocking high-quality AI-ready data as fuel for research, innovation and business models.

Data Actions
Smart AI Governance

Smart AI Governance

Ensure the Swiss way: Innovation-friendly, streamlined AI governance

Governance Actions