Summary

The Cyberconcerns Monitor 2026 by Sotomo opinion research commissioned by AXA shows: Swiss citizens increasingly cannot distinguish AI-generated content from authentic content. Two-thirds of the 1,490 people surveyed in February struggle to recognize AI-generated texts; half find it difficult to distinguish artificial images and audio recordings from real ones. 94 percent demand a labeling requirement for AI content on digital platforms. While the EU mandates platform operators to label content starting in August, the Swiss Federal Council has delayed its new communications platform law for years.

People

Topics

  • Artificial intelligence and labeling requirements
  • Digital media use and screen time
  • Regulation of tech platforms
  • Cybersecurity and consumer trust

Clarus Lead

The study signals a turning point in the digital regulation debate: The Swiss population largely rejects the previous self-regulation strategy and demands state action – particularly regarding AI transparency. This pressure emerges at a moment when the European Union is already setting binding standards, while Bern plays catch-up politically. Simultaneously, the data reveals a paradox: although users want to reduce their screen time themselves, they are massively failing to implement this – an indicator of systemic, not individual problems.

Detailed Summary

The Sotomo study documents a deep trust vacuum. While two-thirds cannot reliably identify AI-generated texts and half are uncertain about images and audio recordings, expectations of those responsible have shifted fundamentally: only somewhat over one-third trusts in user self-responsibility and tech corporations' accountability. The overwhelming majority instead expects tech giants, AI developers, and politicians to intervene regulatorily.

Moritz Büchi from ZHAW clarifies the problem: individual measures such as app avoidance or time limits are technically superficial and fall short. Instead, society must fundamentally clarify what role social media should play – a question that requires state regulation. What surprises Büchi is another finding: more than two-thirds reject personalized content in principle, although personalization is often functional (film, music, or contact suggestions). The media expert suspects a tipping point at scroll algorithms, especially since social media triggers negative feelings for the majority – while music and podcasts more often delight four out of five users.

A third phenomenon reveals psychological dependence: approximately 85 percent of 18- to 35-year-olds habitually reach for their smartphone (in queues, during breaks, on the toilet). "This figure is high; we should keep an eye on it from a societal perspective," warns Büchi. The conclusion remains uncomfortable: Swiss people are aware of their overuse but cannot correct it themselves.

Key Points

  • 94 percent demand a labeling requirement for AI content; the EU implements this starting in August; Switzerland is hesitant
  • Trust in self-regulation collapses: Only one-third still relies on user self-responsibility and platform accountability
  • Systemic addiction mechanisms override self-discipline: 85% of young adults automatically reach for their smartphone; over one-third fails at self-limitation attempts

Critical Questions

  1. Quality of Evidence: How representative are 1,490 respondents (February 2026) for the entire Swiss population? Were age groups, linguistic regions, and socioeconomic differences adequately weighted?

  2. Conflict of Interest: AXA financed the study – to what extent could insurance interests (cybersecurity products, risk perception) have influenced question formulation or interpretation?

  3. Causality: Do Swiss people demand labeling requirements because they cannot recognize AI (competency deficit) or because they fundamentally distrust AI (trust issue)? The study does not differentiate.

  4. Implementation Risks: If platforms must label AI content – who defines the boundaries (spell-checker? statistical tools? fully generated content)? Does strict labeling create new compliance burdens for smaller providers?

  5. Self-Limitation vs. Design: The study shows users fail to reduce their screen time. Is this a willpower deficit or a consequence of targeted algorithm manipulation? Which causal direction applies?

  6. Regulatory Gap: The Federal Council delays the communications platform law – are the reasons political (lobbying) or substantive (technical complexity)? The study does not name them.

Source Directory

Primary Source: Cyberconcerns Monitor 2026 – Sotomo/AXA | Blick Article: https://www.blick.ch/schweiz/kennzeichnungspflicht-gefordert-schweizer-fuehlen-sich-ki-ausgeliefert-id21881219.html

Verification Status: ✓ February 2026

This text was created with the assistance of an AI model. Editorial Responsibility: clarus.news | Fact-Checking: February 2026