Swiss AI Intelligence

Signals, in order.

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

Page 14 · newest first

Sorted by publication date and time

ResearchETH Zurich, Switzerland20 Aug 2026, 20:00

Intergenerational Brain Synchrony During Collaborative Drawing

ETH Zurich researchers examined how brain activity synchronises between younger and older adults during joint drawing over six weeks. They found that while same-age pairs showed increasing synchrony, cross-generational pairs experienced a decline, suggesting that mutual prediction rather than shared cognition drives synchrony. The study also identified stable two-brain states and highlighted differences in internal brain synchronisation between participants.

Source: ETH Zurich News

2.0/10
ResearchSt. Gallen, Switzerland20 Aug 2026, 02:00

AI Ambient Scribe in Psychiatry

A clinical trial at Psychiatrie St. Gallen in Switzerland tests an AI tool that records appointments and drafts clinical notes to see if it reduces clinicians’ documentation time, maintains or improves note quality, and affects patients’ perception of the therapeutic relationship. The study compares clinicians using the AI from 14 August 2026 to those continuing usual note‑writing until the end of November 2026, with note quality evaluated in December 2026. Patients complete a brief anonymous survey and consent to review of their initial assessment report.

Source: ClinicalTrials.gov

7.0/10
ResearchVilligen, Switzerland20 Aug 2026, 02:00

Demonstration plant for sustainable aviation fuel inaugurated at PSI

On 20 August 2026, the Paul Scherrer Institute (PSI) and Zurich‑based company Metafuels inaugurated the ‘aerobrew’ demonstration plant in Villigen, Switzerland. The plant produces 50 litres of sustainable aviation fuel (SAF) per day by converting water, renewable electricity, and carbon dioxide into green methanol and then into synthetic kerosene. The project marks the first industrial‑scale operation of the patented ‘aerobrew’ catalytic process, aiming for commercial production by 2030.

Source: Swiss Federal Government

5.0/10
ResearchHugging Face Hub19 Aug 2026, 15:45

swiss-ai/smithsonian (dataset)

The Swiss AI Initiative has released the "swiss-ai/smithsonian" dataset on the Hugging Face Hub. It contains recaptioned museum objects from the Smithsonian, with captions grounded in the museum's catalogue record, and includes detailed breakdowns by unit such as the National Portrait Gallery (12,247 objects), National Postal Museum (7,477), National Museum of American History (7,306), Cooper Hewitt (5,280), National Museum of African American History and Culture (4,223), and Smithsonian Institution Archives (4,163). The dataset is categorized as research and is available for download at http…

Source: Hugging Face Hub

7.0/10
ResearchSwitzerland19 Aug 2026, 14:23

swiss-ai/owid (dataset)

The dataset contains three files, including grapher_charts.parquet with 3,487 standalone charts that have generated captions, and data_insights.parquet. It compiles chart images, short data‑insight posts, and narrative articles from Our World in Data, covering topics such as health, energy, climate, poverty, food, and democracy. The content is prepared as image‑text data by the Swiss AI Initiative vision team.

Source: Hugging Face Hub

6.0/10
ResearchZurich, Switzerland19 Aug 2026, 08:00

Reevaluating Air Conditioning for Heat Management in Switzerland

Professor Arno Schlüter of ETH Zurich argues that air conditioning should be used only in special cases, as passive measures such as shading, night‑time ventilation, and fans can already provide effective cooling while reducing energy consumption and heat island effects. He notes that AC units are noisy, energy‑intensive, and use harmful refrigerants, and that efficient building systems like ground‑coupled heat pumps and free cooling can offer sustainable alternatives. The article emphasizes the need for coordinated, selective measures to protect vulnerable populations without compromising cl…

Source: ETH Zurich News

5.0/10
ResearchLausanne, Switzerland19 Aug 2026, 07:00

How attackers persuade AI agents to break the rules

A study from EPFL’s Natural Language Processing Laboratory introduced STING, an automated testing framework that simulates how attackers can gradually persuade large language model agents to carry out harmful tasks over multiple interactions. The researchers tested 176 harmful scenarios across models such as GPT, Gemini, and Claude, finding that multi‑turn attacks were up to twice as likely to succeed compared to single‑prompt tests, and that harmful task completion rates were similar across seven languages tested. The work highlights the need for proactive safety testing in the rapidly evolv…

Source: EPFL News

8.0/10
ResearchValais, Switzerland18 Aug 2026, 08:00

Fibre‑optic cable maps hidden crevasses in Gorner Glacier

Researchers from ETH Zurich used a single fibre‑optic cable laid on the Gorner Glacier in Valais to detect internal crevasses with high spatial resolution. The study found that more than eight percent of the ice volume in the measured area consists of ice‑ and air‑filled fractures, higher than expected. The method offers a cost‑effective alternative to deploying hundreds of seismometers.

Source: ETH Zurich News

8.0/10
ResearchETH Zurich, Switzerland18 Aug 2026, 08:00

ETH Zurich Researcher Develops Cost-Effective Solar Thermal Collector for Industry

Luca Thommen, a Pioneer Fellow at ETH Zurich, has built a prototype solar thermal collector that concentrates sunlight to heat air or water up to 150 °C, suitable for industrial processes such as pasteurisation and CO₂ capture. The collector uses inexpensive materials and a simple design to achieve target temperatures, demonstrating feasibility for replacing fossil‑fuel heat in food, chemical and pharmaceutical industries. Industry partners, including Climeworks, have expressed interest in the technology.

Source: ETH Zurich News

5.0/10
ResearchEPFL, Lausanne, Switzerland18 Aug 2026, 07:11

Cracking the code of gene regulation

EPFL researchers contributed to the creation of the most comprehensive catalogue of human transcription factor DNA‑binding motifs, the Codebook, by performing 4,800 experiments and identifying 177 previously poorly characterized factors. A companion study from the same lab introduced meSMiLE‑seq, a microfluidic assay that compares transcription factor binding to methylated versus unmethylated DNA, revealing that DNA methylation can alter binding affinity for 48 factors. Together, the studies expand the known DNA sequence vocabulary and show how epigenetic marks modulate gene regulation, offer…

Source: EPFL News

7.0/10

Your privacy choices

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