This scoping review directs attention to artificial intelligence–mediated informal language learning (AI-ILL), defined as autonomous, self-directed, out-of-class second and foreign language (L2) learning practices involving AI tools. Through analysis of 65 empirical studies published up to mid-April 2025, it maps the landscape of this emerging field and identifies the key antecedents and outcomes. Findings revealed a nascent field characterized by exponential growth following ChatGPT’s release, geographical concentration in East Asia, methodological dominance of cross-sectional designs, and limited theoretical foundations. Analysis also demonstrated that learners’ AI-mediated informal learning practices are influenced by cognitive, affective, and sociocontextual factors, while producing significant benefits across linguistic, affective, and cognitive dimensions, particularly enhanced speaking proficiency and reduced communication anxiety. This review situates AI-ILL as an evolving subfield within intelligent CALL and suggests important directions for future research to understand the potential of constantly emerging AI technologies in supporting autonomous L2 development beyond the classroom.

The impact of misalignment errors, specifically yaw and pitch deviations, on dihedral reflectors’ scattering responses is studied for millimeter-wave polarimetric multiple-input multiple-output automotive radars. Through simulations and experiments at 77 GHz, it is demonstrated that significant radar cross-section (RCS) variations of up to 30 dB can occur within small misalignment ranges (0$^{\circ}$–2$^{\circ}$). The findings emphasize that larger dihedral dimensions can amplify sensitivity to misalignment in some specific misalignment scenarios, offering trade-offs between reflection strength and robustness to misalignment errors. The study also explores near-field effects, revealing notable discrepancies between the dihedral near- and far-field scattering response in misalignment scenarios. A polarimetric calibration method is applied to show how polarimetric channel phase response is affected under such conditions, achieving stable results in specific configurations (e.g., dihedral at 0$^{\circ}$ under yaw misalignment angle). This study addresses key challenges in calibration accuracy, including the high sensitivity of RCS to small angular misalignments, the trade-offs between reflector dimensions and robustness, and the influence of near-field effects in practical setups.

Developments in military AI highlight that maintaining state control over military innovation that is driven by private corporate experts is challenging. Even the leading military AI power, the United States, has struggled to meet this challenge, while trying different modes of control over time. What explains these struggles and the shifting modes of state control over private military innovation? Bringing together the ‘competence–control theory’ of indirect governance and research on technology-driven transformations in the making of national security, we propose a novel theory of state control over military innovation. We argue that states face a trade-off between (fostering) the competence of private corporate experts and (enhancing) state control over military innovation. This trade-off is shaped by technological change and geopolitical competition. Depending on the relative strength of these drivers, varying prioritisations of competence or control lead to different hierarchical or non-hierarchical, capacities- or rules-based modes of control. Tracing the evolution of the US national security state and its relations to private corporate experts in the subfield of autonomous weapons systems, we demonstrate that our theoretical argument explains otherwise puzzling intertemporal variation in control modes. Our findings have important policy implications for the institutional design of military innovation.

Artificial Intelligence (AI) is increasingly being used in the agrifood sector to improve efficiency, safety and sustainability. The application of the European Union Artificial Intelligence Act’s (AI Act) to the broad spectrum of agrifood AI systems has not been thoroughly analyzed. Therefore, this research article explores the AI Act’s application to the agrifood sector by providing an overview of AI systems in the sector and applying the AI Act’s risk-categorisation system to them. Section II provides an outline of the AI Act and its risk categorisation system. Section III gives an overview of agrifood AI systems and explains how the AI Act applies to them. The AI Act is most significantly applicable to AI safety components that are used in agricultural or agroforestry vehicles and other machinery. Although not many, there are also some types of agrifood AI systems that will be prohibited and that will have special transparency requirements under the AI Act. The paper also identifies potential sector-specific challenges such as a risk to food security, an open question regarding the extensiveness of the meaning of ‘AI safety component’ and the inattention of the AI Act to non-human centred values such as the environment, biodiversity and animal welfare.

This study presents data on helminth communities from 93 Hooded Crows (Corvus cornix). The dataset includes historical and contemporary records from three localities in Ukraine with different levels of urbanisation: Kyiv, the Middle Dnipro River, and Polissya. Thirty-two helminth species were identified, including 14 trematodes, six cestodes, 11 nematodes, and one acanthocephalan. The nematodes Eufilariella delicata and Hadjelia truncata are documented in Hooded Crows for the first time. During the statistical analysis, it was revealed that the used dataset is insufficient for robust inference regarding the impact of urbanisation on helminth communities due to its temporal and spatial biases. Despite the limitation, the data offer information for future research on the influence of urbanisation on helminth biodiversity in avian hosts.

We explore the Weihrauch degree of the problems “find a bad sequence in a non-well quasi order” ($\mathsf {BS}$) and “find a descending sequence in an ill-founded linear order” ($\mathsf {DS}$). We prove that $\mathsf {DS}$ is strictly Weihrauch reducible to $\mathsf {BS}$, correcting our mistaken claim in [18]. This is done by separating their respective first-order parts. On the other hand, we show that $\mathsf {BS}$ and $\mathsf {DS}$ have the same finitary and deterministic parts, confirming that $\mathsf {BS}$ and $\mathsf {DS}$ have very similar uniform computational strength. We prove that König’s lemma $\mathsf {KL}$ and the problem $\mathsf {wList}_{{2^{\mathbb {N}}},\leq \omega }$ of enumerating a given non-empty countable closed subset of ${2^{\mathbb {N}}}$ are not Weihrauch reducible to $\mathsf {DS}$ or $\mathsf {BS}$, resolving two main open questions raised in [18]. We also answer the question, raised in [12], on the existence of a “parallel quotient” operator, and study the behavior of $\mathsf {BS}$ and $\mathsf {DS}$ under the quotient with some known problems.

Food insecurity is a global issue. The objective is to summarise the literature identifying the main outcomes related to out-of-school hours interventions that provide food for low-income families with school-aged children, how they impact school-aged children and their families, and to identify gaps in knowledge. This review covered the main types and dimensions proposed in the literature. One author independently selected the studies, and an independent reviewer randomly reviewed them. Any paper meeting the inclusion criteria was considered regardless of geographical location. Papers were predominantly from the US, UK and Australia, including school-aged children from low-income families. Ninety-four articles were included relating to holiday clubs (n = 38), breakfast clubs (n = 45) and after-school clubs (n = 11). Key outcomes were healthy eating, academic, social, physical activity, nutritional education and financial outcomes. Clubs were consistent regarding the positive social and financial outcomes. There was variation in the primary aim, either to improve healthy eating or to feed children, regardless of nutritional quality. None of the studies reported children’s health outcomes. This review identified the key outcomes of interventions for low-income families outside of school hours in the literature. It highlights the consistent positive social outcomes across the three intervention types and the discrepancy in the nutritional value of the food provided. Few studies examined the attainment impact of holiday clubs, with no evidence on how they could impact term-time attendance. This highlights the need to analyse secondary data to understand further the attainment and attendance impact on children attending these interventions over time.

Non-adherence and even partial adherence to antipsychotic treatment can increase the risk of relapse in patients with schizophrenia. One strategy to improve adherence is through the use of long-acting injectable (LAI) antipsychotics. Multiple LAI antipsychotic options are available, which differ in terms of their formulation, administration, initiation, and maintenance dosing schedule. This article provides a practical guide to the conversion from oral to LAI antipsychotic treatment for the available LAI formulations as well as evidence-based principles for maintenance treatment.

EU law lacks a coherent legal framework that adequately defines, protects, and empowers cybersecurity researchers, particularly those operating outside formal institutions. This paper examines how cybersecurity research fits within the evolving EU regulatory landscape, with a particular focus on the Digital Services Act (DSA), the Cyber Resilience Act (CRA), and the NIS2 Directive. It explores the legal ambiguity surrounding researcher status, the conditions for data access and auditing under the DSA, and the challenges posed by current vetting requirements. Drawing on doctrinal legal analysis and interdisciplinary insights from cybersecurity and platform governance, the paper argues that while the DSA provides novel tools such as vetted researcher access and auditing obligations for Very Large Online Platforms (VLOPs) its structure is better suited for systemic risk research than for adversarial, exploratory cybersecurity testing. The paper concludes that a sustainable model for cybersecurity research governance in the EU must go beyond DSA-style vetting, incorporating flexible mechanisms like coordinated vulnerability disclosure and bug bounty programs, as reflected more directly in the CRA.

After implementation of a molecular syndromic panel for infectious diarrhea, a significantly greater proportion of C. difficile results were classified as colonization rather than infection compared to the pre-implementation period. Routine C. difficile reporting from multiplex panels should be re-evaluated to minimize diagnostic uncertainty in some patients.