Key takeaways:

Researchers demonstrate that misleading text in the real-world environment can hijack the decision-making of embodied AI systems without hacking their software.

Self-driving cars, autonomous robots and drones, and other AI systems that use cameras may be vulnerable to these attacks. 

The study presents the first academic exploration of environmental indirect prompt injection attacks against embodied AI systems.

Amidst the accelerated restructuring of the global economic landscape, technological innovation, and industrial transformation, vocational education, as the core carrier of human capital cultivation, is undergoing systematic changes. Between 2022 and 2024, countries have gradually formed a dominant trend toward the modernization of vocational education through policy innovation and practical exploration. Against this backdrop, the landscape of global vocational education is undergoing a transformative shift, driven by technological advancements, economic changes, and evolving workforce needs.

Scientists at the German Cancer Research Center (DKFZ), Heidelberg University Hospital (UKHD), and Mannheim University Medical Center (UMM) are presenting a method that enables artificial intelligence (AI) to learn how to transfer medical image data from animals to humans. This “xeno-learning” could help make surgical procedures safer and more precise in the future—without relying on human training data.

Artificial intelligence is rapidly changing our society and economy. A new study shows that the majority of people believe that artificial intelligence is displacing more human labour than it is creating new opportunities. Scientists at the University of Vienna and Ludwig-Maximilians-Universität München (LMU) demonstrated a causal link: the stronger this perception, the more dissatisfied people are with democracy – and the less they participate in political debates about future technological developments. These effects occur even though artificial intelligence has had only a limited impact on the labour market so far. The study was recently published in the renowned journal PNAS.

The Education University of Hong Kong (EdUHK) has long been committed to advancing education technology that integrates innovation with human-centred design. Yesterday, the University announced the application of two inventions developed by its researchers. First is a language‑learning social robot designed for nursery and primary children. In trials, the robot boosted English comprehension scores by up to 21.2% and extended attention spans by 5.43% during a 20-minutes reading session, outperforming both tablets and more mechanical-looking robots. 

 

The second innovation, EmoCare, is an AI-powered emotional health support application built on the positive psychology model PERMA. It helps users avoid excessive immersion in negative states by embedding positive elements into generated images and conversations, thereby supporting emotional regulation. The launch comes amid growing concern over adolescent suicides linked to unmoderated online chat rooms powered by generative AI without structured psychological safeguards.