Summary

The National Research Program NFP 77 (2020–2025) examined the effects of digitalization on Swiss society, the job market, and democracy across 46 projects. Computer science professor Abraham Bernstein, president of the steering committee, warns: Switzerland must not allow itself to be driven by digital transformation, but must actively shape it. The studies reveal both opportunities (better job placement, digital participation) and risks (news deprivation affecting 48% of the population, gender and age gaps in digital skills).

People

  • Abraham Bernstein (Computer Science Professor, University of Zurich, President of NFP 77 Steering Committee)
  • Mark Eisenäcker (Research Fellow, University of Zurich; Study on News Deprivation)

Topics

  • Digital skills and continuing education
  • Labor market changes due to automation
  • Democratic participation and digital tools
  • Artificial intelligence: transparency and regulation
  • Gender gap and age divide in digitalization

Clarus Lead

Switzerland stands at a turning point in digital transformation. While research and innovation are internationally leading, 46 studies reveal serious deficits: Nearly half the population consumes no news anymore – a risk for informed voting decisions in direct democracy. At the same time, employers massively underutilize their employees' digital skills. The core message of the research results presented today: Digitalization is social change, not merely a technical problem. Decision-makers in politics, business, and educational institutions must actively regulate and invest – or opportunities such as job matching and participation will be lost.

Detailed Summary

Digitalization as Interaction Between Technology and Society

Understanding digitalization goes beyond mere technology adoption. According to Bernstein, it is a matter of mutual influence: technologies change usage patterns (from newspapers to mobile news feeds, from television to streaming), which in turn creates new expectations for developers. This is visible in media consumption, political communication, and work structures.

Education and Job Market: Lifelong Skills Required

The research calls for promoting digital skills throughout life – not only in schools. Employers and employees must share responsibility. Particularly relevant: job platforms should focus less on job titles and more on skill catalogs. An example from the research: someone trained in watchmaking could apply related skills in dental technology – a potential that remains untapped today. Two studies examined how job seekers and employers can better find each other through competency matching.

Democracy Under Pressure: News Deprivation and Digital Participation

A central risk: 48% of surveyed individuals consume no news anymore (study by Mark Eisenäcker). Fatal for direct democracy – citizens can only vote on the basis of information. Solutions from two pilot projects:

  • Community discussions electronically (University of Bern, Canton of Bern): thousands of citizens discussed online before votes; increased legitimacy of results.
  • Participatory budgeting (University of Fribourg + ETH, City of Aarau): citizens help shape investment decisions digitally; higher sense of participation regardless of outcome.

Counteraction: excite people about politics instead of media – those interested in political processes automatically consume more news.

Artificial Intelligence: Reliability Instead of Trust

AI is often criticized as a "black box." Bernstein distinguishes: trust is a feeling (interpersonal); reliability is a measurable property. This can be solved through transparency – e.g., decision trees on the Titanic dataset show clearly: gender was a survival factor, then class. ChatGPT does not offer this traceability. The consequence: users must develop understanding of limitations – AI makes statistical errors and sometimes invents facts. Medical advice via AI is riskier than a recipe suggestion.

Regulation: The Swiss Way Between Innovation and Safety

While the EU regulates strictly (AI Act), Switzerland chooses more flexible boundary-setting. In parallel, a proprietary language model is being developed to preserve digital sovereignty. Critical point: regulation must be enforced – without consequences it fails (e.g., deepfakes in election campaigns).

Gender Gap and Age Divide

A University of Basel study shows: women rate their digital skills significantly lower than men – and consider continuing education less important. Older population groups participate less. Paradox in universities: female professors rate student ChatGPT use lower than students themselves. Bernstein demands: playful engagement with technology to reduce hesitation – mistakes are normal with statistical machines.

Key Statements

  • Digital transformation is an interaction between technology and social application – not linearly technologically determined.
  • Lifelong digital skills require investments from the state, employers, and individuals; current school education is insufficient.
  • News deprivation (48%) endangers informed voting decisions; digital participation tools can increase legitimacy and news consumption.
  • AI reliability requires transparency and user skepticism, not blind trust; application context is decisive (recipe vs. medication).
  • Gender and age gaps in digital skills and security require targeted low-barrier continuing education and positive framing.
  • Switzerland must actively regulate and build sovereign infrastructure rather than be driven by global platforms.

Critical Questions

(a) Evidence / Data Quality / Source Validity

  1. The study mentions "48% news deprivation" – is this the surveyed sample or an extrapolation to the entire population? What demographic characteristics were controlled?
  2. The 46 projects spanned 5 years (2020–2025): how consistent are findings across this period, during which AI (ChatGPT: Nov. 2022+) has accelerated massively?
  3. The job matching studies mention no participant numbers, success rates, or comparability with existing platforms – how reliable is the recommendation?

(b) Conflicts of Interest / Incentives / Independence

  1. NFP 77 was commissioned by the Federal Council – are recommendations for "more flexible regulation" (vs. EU strictness) structurally biased toward Swiss industry interests?
  2. Who finances the "proprietary language model development"? Is there a conflict of interest between academic research and the private AI sector?

(c) Causality / Alternatives / Counter-hypotheses

  1. News deprivation is attributed to technology switching (streaming, social media) – but could reduced trust in established media or deliberate information avoidance also be factors?
  2. Job matching opportunities are based on pilot projects – are scaling effects and market dynamics taken into account? Could large platforms absorb this instead of democratizing it?

(d) Feasibility / Risks / Side Effects

  1. "Lifelong skills" – how are costs divided between the state, employers, and individuals? Risk: continuing education becomes a burnout factor (Bernstein confirms: "adjusting slides every week").

Sources

Primary Source: SRF Tagesgesprächserie (28.05.2026) – "Abraham Bernstein: Die digitale Transformation mitgestalten" – Podcast Audio

Supplementary Sources (cited in transcript):

  1. NFP 77 – National Research Program "Digital Transformation" (2020–2025)
  2. Stanford Study on AI Research (Ranking by Researchers per Capita)
  3. University of Basel – Study on Gender Gap in Digital Skills Self-Assessment
  4. University of Bern – Pilot Project Electronic Community Discussions (Canton of Bern)
  5. University of Fribourg + ETH – Participatory Budgeting (City of Aarau)

Verification Status: ✓ 30.05.2026

This text was created with the support of an AI model.
Editorial Responsibility: clarus.news | Fact-Check: 30.05.2026