The 28th European Conference on Artificial Intelligence (ECAI-2025) is currently taking place in Bologna, Italy, running from 25-30 October 2025. During the opening ceremony, the winners of the ECAI-2025 and Prestigious Applications of Intelligent Systems (PAIS-2025) outstanding paper awards were announced. Letting AI agents interact in multi-agent applications adds a layer of complexity to the interpretability and prediction of AI outcomes, with profound implications for their trustworthy adoption in research and society. Game theory offers powerful models to capture and interpret strategic interaction among agents, but requires the support of reproducible, standardized and user-friendly IT frameworks to enable comparison and interpretation of results. We describe its implementation and usage, and we employ it to uncover biased outcomes in popular games among AI agents, depending on the employed Large Language Model (LLM) and used language, as well as on the personality trait or strategic knowledge of the agents.

Men with suspected prostate cancer will be able to get a diagnosis from the NHS within a day, under a new trial hailed as a potential game changer for identifying and treating the disease. The 15 hospitals taking part will use AI technology to interpret MRI scans and spot areas of abnormal tissue within minutes, according to NHS England. Scans showing a high-cancer risk will be triaged as priority review for a radiologist and patients will be booked for a same-day biopsy. Around one in eight men will develop prostate cancer in their lives, according to Prostate Cancer UK, with research showing it has overtaken breast cancer as the most commonly diagnosed form of the disease in the UK. But unlike breast cancer, there is currently no national screening programme for prostate cancer.

Amazon is planning major job cuts among its corporate workers as soon as this week, multiple media outlets have reported. The online retail giant plans to lay off as many as 30,000 employees as part of cost-cutting measures led by chief executive Andy Jassy, according to the Wall Street Journal and Reuters. Each cited sources stating the same number of layoffs. Amazon declined to comment when contacted by the BBC. If confirmed, the layoffs could be one of the largest seen in recent months.

Could Ukraine hold a presidential election right now? Will Europe use frozen Russian assets to fund war? How can Ukraine rebuild China ties? 'Ukraine is running out of men, money and time' Russian attacks on Ukraine's southern Zaporizhia killed a 44-year-old man and wounded several others, Governor Ivan Fedorov said on Monday, as the death toll from other assaults on Sunday continued to rise. Ukrainian officials said the attacks on Sunday killed two people in the eastern Donetsk region and a 69-year-old man in the northern Sumy region.

Ever since her death in 1986, it was taken as common knowledge that Elsa Lanchester - who became a horror movie icon by playing the title character in the Bride of Frankenstein - had been cremated and her ashes sprinkled in the ocean. But then Scott Michaels, the founder of Dearly Departed Tours, discovered that her cremated remains were interred in a rose garden under her married name, Elsa Lanchester Laughton. For almost 40 years no one had made the connection - until now, he says. Mr Michaels, 63, is a historian who specialises in the dark side of Hollywood. A go-to for programmes about dead Hollywood celebrities and murder, he has consulted for Quentin Tarantino's Manson murder film Once Upon a Time in Hollywood.

Turing AI Institute boss denies accusations of'toxic internal culture' The Alan Turing Institute Chair has told the BBC there is no substance to a number of serious accusations which rocked the organisation in the summer. In August, whistleblowers accused the charity's leadership of misusing public funds, overseeing a toxic internal culture, and failing to deliver on its mission. They said the Turing Institute, the UK's national body for artificial intelligence (AI), was on the brink of collapse after Peter Kyle, the then technology secretary, threatened to withdraw its £100m funding. But speaking exclusively to the BBC, Chair Dr Doug Gurr said the whistleblower claims were independently investigated by a third party which found them to have no substance. I fully sympathise that going through any transition is always challenging, he said.

These robots can clean, exercise - and care for your elderly parents. Would you trust them to? Hidden away in a lab in north-west London three black metal robotic hands move eerily on an engineering work bench. We're not trying to build Terminator, jokes Rich Walker, director of Shadow Robot, the firm that made them. Bespectacled, with long hair and a beard and moustache, he seems more like a latter-day hippy than a tech whizz, and he is clearly proud as he shows me around his firm.

We evaluate the performance of targeted maximum likelihood estimation (TMLE) for estimating the average treatment effect in missing data scenarios under varying levels of positivity violations. We employ model- and design-based simulations, with the latter using undersmoothed highly adaptive lasso on the 'WASH Benefits Bangladesh' dataset to mimic real-world complexities. Five missingness-directed acyclic graphs are considered, capturing common missing data mechanisms in epidemiological research, particularly in one-point exposure studies. These mechanisms include also not-at-random missingness in the exposure, outcome, and confounders. We compare eight missing data methods in conjunction with TMLE as the analysis method, distinguishing between non-multiple imputation (non-MI) and multiple imputation (MI) approaches. The MI approaches use both parametric and machine-learning models. Results show that non-MI methods, particularly complete cases with TMLE incorporating an outcome-missingness model, exhibit lower bias compared to all other evaluated missing data methods and greater robustness against positivity violations across. In Comparison MI with classification and regression trees (CART) achieve lower root mean squared error, while often maintaining nominal coverage rates. Our findings highlight the trade-offs between bias and coverage, and we recommend using complete cases with TMLE incorporating an outcome-missingness model for bias reduction and MI CART when accurate confidence intervals are the priority.

The ability to accurately align LLMs with human population groups on subjective questions would have great value. In this work, we show that use of simple supervision can greatly improve language model alignment with diverse population groups more consistently, as measured over three datasets spanning various topics. Beyond evaluating average alignment, we also report how alignment varies across specific groups. Our broad findings provide insights into the distributional alignment of LLMs with diverse population groups. By conducting evaluation over many LLMs and prompting strategies, along with open-sourcing our work, we provide a benchmark to stimulate future research.

Pre-trained models exhibit strong generalization to various downstream tasks. However, given the numerous models available in the model hub, identifying the most suitable one by individually fine-tuning is time-consuming. In this paper, we propose \textbf{SwiftTS}, a swift selection framework for time series pre-trained models. To avoid expensive forward propagation through all candidates, SwiftTS adopts a learning-guided approach that leverages historical dataset-model performance pairs across diverse horizons to predict model performance on unseen datasets. It employs a lightweight dual-encoder architecture that embeds time series and candidate models with rich characteristics, computing patchwise compatibility scores between data and model embeddings for efficient selection. To further enhance the generalization across datasets and horizons, we introduce a horizon-adaptive expert composition module that dynamically adjusts expert weights, and the transferable cross-task learning with cross-dataset and cross-horizon task sampling to enhance out-of-distribution (OOD) robustness. Extensive experiments on 14 downstream datasets and 8 pre-trained models demonstrate that SwiftTS achieves state-of-the-art performance in time series pre-trained model selection.