Switzerland prepares AI legislation for spring 2027 consultation

The Swiss Federal Council on 28 October 2026 took note of interim progress on implementing the Council of Europe Convention on Artificial Intelligence. The Federal Office of Justice (FOJ) is preparing legally binding measures, with a consultation scheduled for spring 2027. The Federal Office of Communications (OFCOM) is coordinating the parallel non-binding work, including guidelines, industry standards and voluntary commitments.

For compliance teams, this is a policy milestone rather than a new legal duty. The Federal Council’s regulatory approach envisages sector-specific legislative amendments wherever possible, with general rules reserved for key fundamental-rights areas such as data protection. The future measures are intended to address transparency, data protection, non-discrimination, supervision and risk assessment.

Two tracks for implementation

The FOJ is working with OFCOM and the Federal Department of Foreign Affairs to prepare the draft act for consultation.

That draft remains at the preparation stage. The spring 2027 consultation will therefore be the next point at which firms can assess the proposed duties and scope.

The non-binding strand is further advanced. OFCOM’s report on sector-specific AI measures is based on a survey of all federal agencies and covers current and planned activity in areas including transport, energy, health and media. It identifies particular challenges for high-tech sectors, intellectual property, data processing, digital infrastructure and regulated markets.

Atlas reads the structure as a deliberate alternative to a single horizontal AI statute: the likely compliance impact will depend on a firm’s sector and the later legislative amendments, while cross-sector principles are being developed through the Convention and general fundamental-rights rules.

OFCOM publishes a voluntary-code toolkit

OFCOM has also published its *Guide on an AI Code of Conduct*, a toolkit for associations and businesses. It contains suggested wording on 10 topics at three levels of detail, which users can combine, and sets out a staged process from drafting through implementation, review and continuous improvement.

The guide is voluntary guidance, not an enforceable code. Firms can use it to structure internal or sectoral commitments, but the Federal Council has not turned those suggested texts into mandatory controls.

Industry dialogues are already producing sector-specific work:

- The Federal Office of Culture has run an AI dialogue with the creative sector since November 2025, including work towards a code of ethics.
- The Federal Office of Energy has asked the electricity sector to develop industry guidelines and voluntary commitments addressing AI risks to electricity supply.
- OFCOM held discussions with the media sector in May on a voluntary commitment.

The Federal Council also tasked the Department of the Environment, Transport, Energy and Communications, through OFCOM, with creating an internal administrative knowledge-sharing platform on sectoral AI self-regulation by the end of March 2028. It is also expected to provide methodological support to departments and further develop the AI code-of-practice guidance.

The immediate compliance task is therefore monitoring rather than implementation of binding rules: firms in affected sectors should follow the relevant dialogue and consider whether the toolkit can support their existing AI governance.