Understanding AI Transparency Requirements: What Swiss SMEs Need to Know

AI Transparency: Why It Matters Now for Swiss SMEs

With EU transparency obligations for generative AI now active, Swiss SMEs must act to ensure users understand when and how they’re interacting with AI-driven systems. Transparency isn’t just about legal compliance; it’s about building trust, clarifying responsibility, and maintaining a competitive edge in an increasingly regulated digital economy.

What Is AI Transparency?

At its core, AI transparency means making it clear to users, customers, and regulators when they are dealing with artificial intelligence—especially generative AI systems that can produce human-like text, images, or decisions. Transparency also covers the disclosure of important information about how these systems work, their limitations, and their data sources, where required by law or best practice.

Key aspects of AI transparency include:

• Notifying users when AI-generated content or interactions occur
• Disclosing the capabilities and limits of the AI system
• Providing accessible documentation about data usage and model behavior
• Allowing users to seek human intervention when needed

New EU AI Act Requirements (August 2026)

Although many high-risk provisions of the EU AI Act are delayed until 2027 or later, core transparency rules for generative AI are now in effect. These rules impact Swiss SMEs if they provide AI-powered services to EU users or process EU citizens’ data. Even if your business is based in Switzerland, the so-called “Brussels effect” means these obligations often apply extraterritorially.

The main requirements include:

• AI Disclosure: Clearly inform users whenever they are interacting with an AI system—especially if the AI can produce content indistinguishable from human output (e.g., chatbots, automated emails, content generation tools).
• Output Labelling: For generative AI, mark AI-created images, audio, or text—such as with watermarks, metadata, or explicit statements—so users do not confuse them with human-produced content.
• Transparency Information: Provide concise documentation explaining what your AI system does, its intended use, known limitations, and how users can challenge or appeal outcomes.

Why Does Transparency Matter for Swiss SMEs?

Implementing transparency is not just a regulatory formality. For Swiss SMEs, it is an essential step for:

• Earning Customer Trust: Customers, especially in regulated industries or public services, value clarity about when AI is involved.
• Reducing Legal Risk: Early compliance helps avoid fines or reputational damage, especially when serving EU clients.
• Improving Internal Governance: Transparency documentation supports better understanding and control over AI deployments inside your company.
• Market Differentiation: Businesses that are open about their use of AI can position themselves as responsible actors in a crowded marketplace.

How to Implement AI Transparency Practically

Here’s how a typical Swiss SME can meet the new transparency requirements:

1. Map Your AI Use Cases

• List all customer- or user-facing systems where AI is involved (chatbots, automated document generation, recommendation engines, etc.).
• Identify which of these systems interact with EU users or process EU data.

2. Add User Notifications

• Place clear statements or icons indicating when users are communicating with AI (e.g., “This response was generated by an AI assistant”).
• For images or audio, consider visible or metadata-based watermarks.

3. Document AI Model Information

• Prepare a simple, accessible factsheet for each AI service, covering: purpose, data sources, basic functioning, known limitations, and escalation contact for human review.
• Make this documentation available on your website or within product interfaces.

4. Train Staff and Update Policies

• Brief customer support and compliance teams on new obligations.
• Update privacy policies and user terms to describe AI use transparently.

5. Monitor and Respond

• Regularly review interfaces and documentation for clarity and accuracy.
• Encourage and handle user feedback about AI-generated content or automated interactions.

Swiss-Specific Considerations

Swiss SMEs face the added complexity of dual compliance: Swiss law (like the revised Data Protection Act) already values transparency, but EU rules may go further. Adopting the stricter standard is often the safest approach, especially if you serve international clients.

For SMEs using open-source Swiss AI such as Apertus, transparent documentation is often easier to provide, since source code and model details are public. When using proprietary or third-party models, work closely with your suppliers to obtain and share necessary information.

Practical Example

Consider a Zurich-based SME offering a SaaS contract analysis platform for Swiss and EU users. Under the new rules, the company must:

• Clearly inform users when contract summaries or draft clauses are AI-generated.
• Label downloadable documents created by the AI with a note or watermark.
• Publish a clear factsheet describing the AI’s function, data training sources (if known), and limitations (e.g., “AI-generated summaries should not be used as legal advice without human review”).
• Provide a contact channel for users to request clarification or human intervention.

The Bottom Line

AI transparency is now a legal requirement and a business imperative. By proactively implementing user notifications, model factsheets, and clear labelling, Swiss SMEs can not only ensure compliance, but also build credibility and win customer trust in a rapidly evolving AI landscape.