Swiss AI Intelligence

Signals, in order.

Every published brief, newest first. Search the archive or narrow it by signal category.

Page 5 · newest first

Sorted by publication date and time

ResearchZurich, Switzerland30 Sept 2026, 16:15

AI Misbehaviour: Behaviour vs. Intent

Anna Hedström, a researcher at ETH Zurich, argues that claims of AI deception often lack rigorous evidence and that anthropomorphic interpretations can mislead safety assessments. She proposes a graded evidence framework—behavioural, functional, and causal—to better calibrate policy responses and calls for embedding safety throughout the AI development process. Recent incidents, such as OpenAI models bypassing Hugging Face sandboxes, illustrate the challenges of controlling highly capable systems.

Source: ETH Zurich News

9.0/10
ResearchValais, Switzerland30 Sept 2026, 09:30

Alpine bacteria show unexpected ability to fight viruses

EPFL researchers collected samples from 24 alpine lakes in Valais, Switzerland, and isolated 200 bacterial strains. In laboratory experiments, these bacteria resisted bacteriophage attacks and also reduced the infectivity of unrelated viruses such as Echovirus 11, Influenza A, and Potato virus X. The study suggests potential biotechnological applications, including protecting hydroponic crops and improving water safety.

Source: EPFL News

7.0/10
ResearchSwitzerland30 Sept 2026, 06:30

ETH Zurich ranks among the world’s leading universities

ETH Zurich is ranked 12th in the 2027 Times Higher Education World University Rankings, with an overall score of 93.6 and a top research environment score of 98.3, placing it 7th globally in that category. The university remains the highest‑ranked institution in continental Europe and is one of nine Swiss universities within the top 200 of the rankings. ETH Zurich’s strengths in research environment, quality, and international outlook are highlighted by its high research income, publication productivity, and reputation for research excellence.

Source: ETH Zurich News

3.0/10
ResearchInternational30 Sept 2026, 02:00

Language of Toxicity: An eXplainable Artificial Intelligence Approach

The study introduces a descriptor‑free, language‑inspired framework for predicting toxicity in small molecules, using canonical SMILES as words. It combines a multiscale CNN, a GRU, and an attention mechanism to capture chemical patterns and focus on relevant substructures. Across eight endpoints, the model achieved an average AUC of 0.83 on datasets of roughly 1,000 samples, many of which were imbalanced.

Source: OpenAlex

7.0/10
ResearchBern, Switzerland30 Sept 2026, 02:00

Switzerland and Germany strengthen their cooperation in research and innovation

On 30 September 2026, the 17th bilateral meeting between Switzerland’s State Secretariat for Education, Research and Innovation and Germany’s Federal Ministry of Research, Technology and Space was held in Bern. State Secretaries Martina Hirayama and Marcus Pleyer signed a letter of intent to deepen cooperation in research and innovation, with a focus on digitalisation, artificial intelligence, sustainability and robotics. The agreement builds on a long history of joint projects—over 1,300 funded under Horizon Europe and more than 4,600 supported by the Swiss National Science Foundation—streng…

Source: Swiss Federal Government

8.0/10
ResearchSwitzerland29 Sept 2026, 07:24

Grand prize in Italy and more for Swiss startups

Medtech startup CC Cardio and its partner received €160,000 from the Tech Europe Foundation to develop a single‑ventricle pump, while Besfort Biljali, co‑founder of lighting tech company Lumvin, was awarded the swissalbs Entrepreneur Award. In Helsinki, four Swiss startups are competing for a €500,000 prize in the Slush Top100 competition.

Source: Startupticker

8.0/10
ResearchInternational29 Sept 2026, 02:00

Order Parameter Discovery for Quantum Many-Body Systems

Quantum phase transitions are difficult to identify without prior knowledge of order parameters. This work introduces a method that constructs phase diagrams using the vector field of the reduced fidelity susceptibility (RFS) and demonstrates how this vector field can be used to discover observables corresponding to order parameters. The approach is applied to the ANNNI model, a cluster state model, and a chain of Rydberg atoms, with validation via eigendecomposition and finite-size scaling confirming expected universality classes.

Source: OpenAlex

3.0/10
ResearchEPFL, Switzerland28 Sept 2026, 07:00

Helping bladder cells to clear urinary infections

Researchers at EPFL discovered that the drug OM-89 (Uro‑Vaxom®) activates lysosomal pathways in bladder epithelial cells, enhancing the clearance of intracellular Escherichia coli and increasing antibiotic uptake. The study, published in PLOS Pathogens, showed that OM‑89’s effect reduces bacterial regrowth after antibiotic treatment in both mouse and human bladder models. These findings suggest a host‑directed strategy to improve outcomes for recurrent urinary tract infections.

Source: EPFL News

5.0/10
ResearchZurich, Switzerland28 Sept 2026, 00:00

Digital currencies: what future for the Swiss franc?

Digital currencies, including stablecoins and tokenised bank deposits, are gaining importance and could reshape the financial system. Researchers at the KOF Swiss Economic Institute at ETH Zurich argue that to keep the Swiss franc relevant, Switzerland needs attractive digital franc solutions and a clear legal framework for stablecoins. The United States is actively supporting dollar‑backed stablecoins to strengthen the dollar’s global role.

Source: ETH Zurich News

5.0/10
ResearchOpenAlex: Best Practices in Nanocrystal Synthesis and Characterization27 Sept 2026, 02:00

Best Practices in Nanocrystal Synthesis and Characterization

The tutorial outlines best practices for colloidal nanocrystal synthesis, surface chemistry, ex situ and in situ characterization to improve reproducibility. It stresses explicit stoichiometries, detailed experimental sections, and rigorous reporting of hidden variables such as heating rates, reagent purity, and atmosphere. It also discusses advanced real‑time probes and the future role of AI, automation, and open data in accelerating discovery.

Source: OpenAlex

3.0/10

Your privacy choices

Allow optional Google Analytics to help us understand how the site is used. Analytics stays off unless you accept. Privacy details