Introduction: The expanding complexity of dental diseases has exposed the limitations of conventional heuristic-based therapeutic planning. Artificial Intelligence (AI) has evolved beyond diagnostic assistance to become a powerful tool in therapeutic decision-making, enabling data-driven, predictive, and personalized dental care. Material and Methods: This narrative review critically evaluates contemporary applications of artificial intelligence in therapeutic planning, prognostic assessment, and surgical execution across major dental specialties, including periodontology, endodontics, prosthodontics, orthodontics, and implantology. Evidence from machine learning, deep learning, computer vision, and robotics-based systems was synthesized to assess clinical relevance beyond diagnostic accuracy. Results: Artificial intelligence-based clinical decision support systems demonstrated improved precision in treatment planning, outcome prediction, and procedural execution. Applications such as generative prosthetic design, Artificial intelligence -guided endodontic access, implant navigation, orthodontic treatment simulation, and robotic-assisted surgery showed potential to reduce operator variability and enhance therapeutic outcomes. However, challenges related to data heterogeneity, algorithmic bias, explainability, and medico-legal accountability persist. Conclusion: Artificial Intelligence is redefining therapeutic decision-making in dentistry by augmenting clinical judgment rather than replacing it. When integrated within a human-in-the-loop framework, artificial intelligence serves as a high-level therapeutic assistant capable of improving accuracy, efficiency, and personalization of dental care. Future research must prioritize longitudinal clinical validation and ethical governance to enable safe and effective clinical translation.

Towards extreme application scenarios: perspectives on artificial intelligence-driven smart energy management systems

Glucose is the primary energy substrate for the human fetus, essential for brain development and overall growth. Traditionally, fetal glucose supply has been attributed to the maternal-fetal glucose gradient. However, emerging evidence indicates that the placenta plays an active role in regulating glucose availability through its own metabolic processes. This brief review aims to synthesize current knowledge on placental glucose handling in uncomplicated human pregnancies, emphasizing mechanisms beyond passive transfer and quantitative aspects. The placenta exhibits dynamic glucose metabolism, including consumption, storage, and endogenous production. Glycogen pools within placental cells are more likely endogenous sources of glucose than gluconeogenesis. Placental endogenous glucose may represent an auxiliary system that safeguards fetal glucose supply during maternal hypoglycemia and/or increased fetal demand. In vivo studies demonstrate that up to 70% of glucose released to the fetus can originate from placental sources at certain time points. Aerobic glycolysis (with lactate production) is a prominent feature of placental metabolism, with substantial lactate export to the maternal circulation. This energy loss is partly compensated for by uptake of maternal ketone bodies and acetate, highlighting the placenta's flexibility in substrate utilization. These adaptations underscore the placenta's dual role: maintaining its own structural and functional integrity while ensuring fetal oxygen and energy needs. Understanding these mechanisms is critical for defining the partition of energy between the placenta and fetus and its implications for fetal growth, particularly under conditions of maternal nutrient restriction. Insights into placental glucose metabolism may inform strategies for understanding and managing growth deviations and guide development of artificial placental systems.

Personal AI refers to artificial intelligence systems designed to assist individuals in their daily lives by providing smart and personalized support. It helps users manage tasks such as scheduling, reminders, emails, and communication more efficiently. Personal AI systems learn from user behaviour and preferences to deliver customized recommendations and solutions. They enhance productivity by automating repetitive and time-consuming activities, allowing individuals to focus on more important work. Common examples include virtual assistants, smart home devices, and personalized mobile applications. Personal AI can also support education by offering tailored learning experiences and instant access to information. In healthcare, it can help monitor fitness, track health data, and provide basic medical guidance. It improves decision-making by analysing data and suggesting better options based on patterns. Privacy and data security remain important concerns, as these systems often rely on personal information. With rapid advancements in technology, personal AI is becoming more accurate, efficient, and widely accessible. It enables seamless interaction through voice, text, and even visual inputs. Personal AI also plays a role in entertainment by recommending music, movies, and content based on user interests. Businesses use personal AI to improve customer experience and engagement. As development continues, it is expected to become an essential part of everyday life. Overall, personal AI aims to make life easier, smarter, and more convenient for individuals.

A lot of work has been carried out in order to develop an artificially intelligent system that thinks similarly to a human being. Some of the abilities that make up such a thinking process include reasoning and decision making. The reasoning process can be broadly classified into two categories, which include deductive reasoning and inductive reasoning. Deductive reasoning entails deriving specific conclusions based on theories and rules while inductive reasoning entails deriving general conclusions based on specific facts. This research paper focuses on defining deductive and inductive reasoning and comparing the differences between the two. The applicability of both deductive and inductive reasoning in artificially intelligent systems is considered in this paper together with a comparative study. Additionally, the advantages and disadvantages of both deductive and inductive reasoning are provided along with their importance in the domain of machine learning. Apart from the aforementioned information, some of the challenges associated with the development of artificial intelligence systems are discussed in this paper. Among other things, algorithmic biases and explainability influence the decision-making capabilities of intelligent machines.

Generative artificial intelligence systems, including large language models (LLMs), are increasingly used in health-related contexts. A documented feature of these systems is the production of hallucinations, outputs that are factually incorrect, fabricated, or misleading. While the World Health Organization (WHO) has described hallucinations as a "risk and challenge," this paper argues that under the WHO’s ethical framework, certain health-related hallucinations should be understood as forms of harm. Drawing on WHO Principle 5.2, requiring protection from mental and physical harm, this paper examines how health-related hallucinations generate psychological distress, enable harmful decision-making, and undermine patient trust. The reclassification of hallucinations from risk to harm carries significance for the regulatory obligations of AI developers, deployers, and policymakers.

Technological advancements driven by engineering have become key factors in elevating Türkiye's global standing in health tourism, greatly strengthening its competitiveness and attracting a growing number of international patients. Emin Taner Elmas is not a dentist or endodontist, but a Mechanical Engineer and academic. His work focuses not directly on classical dental clinical practice or endodontics, but rather on interdisciplinary fields such as thermodynamics, energy transfer, fluid mechanics, and biomedical engineering. However, Elmas's engineering approach has the potential to contribute to the medical field, including dentistry, indirectly through biomedical and health technologies: Biomedical Approach: Treating the human body as a "bio-machine," Elmas develops theories on the natural vibration frequencies of organs and tissues. This "bio-robotic resonance" theory can inspire the design of next-generation devices for tissue healing or disease diagnosis at a theoretical level. Medical Device Modeling: His expertise in thermodynamics and fluid mechanics is used in the design and simulation of medical devices (e.g., hemodialysis machines or drug delivery algorithms). The mechanical strength of surgical instruments used in dentistry or the thermal effects of dental lasers are engineering problems that fall within Elmas's area of ​​expertise. Interdisciplinary Technologies: He has studies on machine learning and artificial intelligence-supported diagnostic systems. These technologies are increasingly used in the field of endodontics today, such as caries detection and root canal anatomy analysis. In summary, Emin Taner Elmas is not a dentist, therefore he does not develop clinical endodontic procedures. However, his work applying engineering principles to the biomedical field has the potential to contribute to the scientific infrastructure of future dental technologies (device design, diagnostic algorithms, etc.). The "Bio-robotic Resonance and Thermodynamic Interaction" theory and medical technology models developed by Emin Taner Elmas can be indirectly adapted to the fields of dentistry and endodontics. The potential contributions of Elmas's work to dental technologies can be evaluated under the following headings: Bio-robotic Resonance and Diagnosis: Elmas views the body as a "bio-machine," arguing that each tissue has its own unique natural vibration frequency. This approach could form the basis for the development of next-generation diagnostic devices that can detect the condition of tooth canals or microcracks in the tooth root using acoustic signal analysis and Fourier transforms in endodontics. Smart Drug Algorithms: His work focuses on smart drug algorithms and simulations via "Frequency Modulation". This modeling can be used to optimize the thermodynamic interaction of disinfectants or drugs applied into the root canal with the tissue in endodontic treatments. Medical Device Modeling: As a thermodynamics and fluid mechanics specialist, Elmas works on the prototype design and simulation of medical devices (such as hemodialysis machines). This engineering knowledge can directly address specific engineering problems in dentistry, such as controlling the thermal effects of dental lasers or increasing the mechanical efficiency of surgical instruments. Interdisciplinary Approach: His work generally focuses on "Medical Technology," combining mechanical engineering and medical sciences. This perspective contributes to the development of the mechanical and software infrastructure of advanced technologies such as digital intraoral scanners and robotic surgical support systems, which are becoming increasingly common in dentistry today. In summary, Elmas's contribution focuses on the engineering design and theoretical physics of smart devices and diagnostic systems used in dentistry, rather than a clinical application.[1-73]

Living systems achieve precise control over macromolecular synthesis within the confined pockets of enzymes. Reproducing such spatial regulation in artificial systems remains a significant challenge in molecular engineering. Although artificial nanoconfinement has been explored to regulate polymerization, tunable and length-specific peptide growth has yet to be realized. Here, we present a highly stable metal-organic framework (MOF) as an artificial oligomerization platform that enables confined N-carboxyanhydride (NCA) ring-opening polymerization (ROP) with dynamically tunable short peptide growth. Reaction kinetics and NMR analyses reveal a confined oligomerization process within one-dimensional channels segmented by Zr6 clusters. Compared with unconfined solution NCA polymerization showing nearly no chain-length selectivity in short-chain growth, these regularly repeating subcompartments provide defined reaction pockets that achieve pronounced chain-length selectivity and remarkably high peptide loading while maintaining framework integrity. The short peptide chain length can be dynamically tuned by monomer size and reaction kinetics, converting confinement from a template limitation into an active design principle. Moreover, pre-coordination of amino acids at Zr6 clusters modifies the pocket microenvironment, accelerating NCA-ROP by over 30-fold. Selecting suitable monomers further enables the formation of A-(B)n sequence peptides. This work establishes a robust MOF-based platform inspired by natural biosynthetic machinery, advances understanding of confined oligomerization and polymerization, and provides a versatile platform for creating functional peptide-MOF materials with potential application in biomimetic catalysis.

Water and proton transport across biological membranes is essential for cellular function. The development of stimuli-responsive artificial transport systems to replace malfunctioning protein channels is crucial for potential therapeutic applications. Here, we report a photoswitchable azobenzene-based channel that envelops a coherent water wire and efficiently regulates water and proton transport across lipid bilayers using light as an external trigger. The molecule undergoes reversible E ↔ Z photoisomerization, where the Z-isomer adopts a compact, folded, π-stacked conformation encapsulating and stabilizing a water wire that can assemble into a well-defined, directional channel. In contrast, the E-isomer forms an extended, less ordered conformation that disrupts coherent water alignment. The synthesised proton/water transporter demonstrates robust photoswitching behavior, maintaining function over multiple irradiation cycles. Ion transport assays reveal complete rejection of alkali metal cations for both isomers, with no detectable ion transport even in the presence of the protonophore FCCP. Proton transport assays in the presence of valinomycin show that the Z-isomer exhibits a 100% increase in electrogenic proton transport rate compared to the E-isomer at higher concentrations. Water permeability measurements mirror this trend, with the Z-isomer showing 20-35% higher net water transport relative to the E-form. These results highlight how photo-controlled conformational changes in foldamer architecture can be leveraged to create efficient, light-regulated water and proton channels.

Membrane reshaping and channel formation often emerge from the coordinated interplay between membranes and their associated proteins. Reconstituting such integrated behaviors in synthetic systems remains challenging, especially when seeking reversible and programmable control. Here we present a protocol for the construction and application of reconfigurable DNA nanorafts that couple membrane morphology modulation with the formation of large, gated membrane channels in synthetic cells. These nanorafts are DNA origami structures that are functionalized with cholesterol anchors and undergo reversible shape transformations. Upon membrane binding, conformational changes drive their self-arrangement into locally ordered domains that deform giant unilamellar vesicles (GUVs). During GUV shape recovery, aided by protein nanopores, the nanorafts interact with the membrane to form synthetic channels capable of transporting large biomolecules (~70 kDa). These channels can be reversibly sealed, allowing programmable control over membrane permeability. Unlike conventional DNA-based nanopores that require pre-assembly and membrane insertion, this protocol supports stepwise, membrane-coupled channel formation with reversible gating. The protocol details the preparation of DNA nanorafts, membrane-bound conformational control, GUV formation, membrane remodeling, morphology-coupled large channel formation and sealing, as well as quantitative fluorescence microscopy assays to analyze vesicle morphology changes and cargo transport. Standard DNA nanotechnology tools and fluorescence microscopy techniques are sufficient to perform the workflow, which can be completed in ~4 d. The system's modularity makes it broadly applicable for constructing artificial cellular systems with programmable structure and function.

The article examines the current state of legal regulation of health technology assessment (HTA) in Ukraine in the context of rapid innovation development of the national healthcare sector. It substantiates the need to perceive state HTA not merely as a procedural mechanism for selecting medical technologies, but as an institutional instrument for implementing public innovation policy. Particular attention is paid to medical technologies based on artificial intelligence, which are considered both an emerging area of HTA and a potential tool for improving assessment processes, in line with European Union regulations and international best practices. The study analyzes inconsistencies between strategic policy documents in the fields of healthcare and innovation and the existing regulatory framework for HTA, including the Cabinet of Ministers of Ukraine Resolution No. 1300. The current HTA procedure remains largely focused on medicinal products and does not fully address digital health solutions, artificial intelligence systems, or other advanced MedTech innovations prioritized by national strategies, though such situation is surely influenced by wartime conditions. The article proposes legislative amendments to align HTA regulation with strategic innovation priorities, including the explicit recognition of digital and AI-based medical technologies as independent objects of assessment. In addition, the feasibility of developing a dedicated HTA guideline for artificial intelligence systems is justified, drawing on the experience of the European Union, NICE (United Kingdom), Digi-HTA (Finland), and CADTH (Canada). The author concludes that, under appropriate regulatory conditions, HTA can serve as an effective mechanism for stimulating innovation while ensuring patient safety, economic efficiency, and a balance between public and private interests in the healthcare system.

To effectively communicate and collaborate with others, we must monitor not only other people's cognitive states (e.g., what someone thinks or believes) but also their metacognitive states (e.g., how confident they are in their beliefs). While humans routinely share confidence, either explicitly (e.g., "I am sure") or implicitly (e.g., via response times), metacognitive capabilities are still developing in artificial intelligence (AI), raising the question of how humans attribute confidence to AI systems. In seven pre-registered experiments (post-exclusion Ns = 113, 109, 56, 59, 52, 60, 57), participants observed human and AI agents make perceptual choices and reported how confident the observed agent seemed in each choice. Overall, attributions of confidence were sensitive to observed behaviour (e.g., task difficulty, accuracy, and response times), but also agent type: observers consistently overestimated the confidence of AI agents compared to humans-even when their behaviour was identical. This illusion of greater confidence in AI decisions was robust across behavioural profiles, agent descriptions, and decision-making domains (visual perception, general knowledge) but was reduced in more subjective decisions (emotion categorisation). An experimental manipulation further showed that illusions of confidence are rooted in prior beliefs about the agents' capabilities. Together, these investigations of metacognitive attributions reveal a powerful illusion of confidence in artificial systems and highlight a central role for attributions of metacognitive states in human-human and human-AI interactions.

The rapid advancement in the field of quantum information processing sparked great interest in developing what are believed to be new promising qubits made of artificial atoms as well as customized natural ones. Those qubits enjoy mutual interaction with other qubits of the same type or different one in what is known as the hybrid quantum systems. In this work, an asymmetrical generalized version of the famous Tavis-Cummings model is used to prototype the behavior of such hybrid systems. The problem is explored in its full generality by analytically solving the model while upholding realistic conditions, namely two mutually interacting non-identical qubits coupled off-resonance to a radiation field, with non-linearity in both the cavity medium and the qubit-field interaction. The work focuses on studying the entanglement dynamics of the system and its asymptotic behavior, with special emphasis on entanglement sudden death (ESD) and birth (ESB). The system is found to exhibit typical death-revival pattern, which asymptotically approaches a pseudo-steady state that is considerably sensitive to the degree of asymmetry between the qubits, the initial state, and other system parameters. We show that for certain initial entangled states, the system parameters can be tuned to provide maximized persistent terminal entanglement values that are steady and free of any collapse-revival patterns or ESD. In contrary, the disentangled initial states asymptotically attain scattered entanglement maxima that lack any systematic behavior. This terminal entanglement behavior can be attained more rapidly by strengthening the mutual interaction between the qubits, a process that is now experimentally accessible with contemporary artificial qubit systems. These results pave the road for preparing strongly entangled qubits in long-life steady states that are crucially needed for performing efficient quantum information algorithms. These systems can be realized using typical qubits such as quantum dots, trapped ions, superconductors, and Rydberg atoms, in either cavity or circuit QED structures.

Free-living amoebae (FLAs) are ubiquitous protozoa found in soil, air, and artificial systems, including hospital environments. Some genera of free-living amoebae, such as Acanthamoeba, can cause severe health complications, including Acanthamoeba keratitis and granulomatous amoebic encephalitis. This study investigated the presence of free-living amoebae (FLAs) in hospital environments, including ready-to-use medical devices and equipment (such as lasers, swabs, and forceps), as well as beds and gowns. To the best of our knowledge, FLAs in these medical devices and equipment have been examined for the first time. In this cross-sectional study, 45 environmental and medical device samples were collected from two hospitals in Northwestern Iran. After filtration, the samples were cultured in a 1.5% non-nutrient agar medium enriched with Escherichia coli. The growth of the FLAs and their genera was determined through microscopic analysis. Positive samples were submitted for PCR analysis targeting the 18S rRNA gene, followed by sequencing and phylogenetic analysis. Also, the pathogenicity of Acanthamoeba isolates was evaluated through osmo- and thermotolerance tests. FLAs were detected in 22.22% (10/45) of samples using microscopy. Most of the examined sources (90%) had mixed contamination, including Acanthamoeba, Vahlkampfia, and Veramoeba (4), Acanthamoeba and Vahlkampfia (1), Acanthamoeba and Veramoeba (2), and Veramoeba and Vahlkampfia (2). Also, one source showed sole contamination with Vahlkampfia. Among the positive samples, 5 were obtained from environmental sources, 4 from equipment, and 1 from surgical gowns. Most Acanthamoeba isolates demonstrated osmo-tolerance (72.48% at 0.5 M) and thermo-tolerance (100% at 37°C).. Sequence analysis identified Acanthamoeba T4 genotype (5), Vahlkampfia sp. (3), and V. vermiformis (6). In this study, FLAs were isolated from patients' beds and surgical gowns for the first time, emphasizing new infection risks within an ophthalmology hospital. In addition to the high prevalence of FLAs in the examined sources, most Acanthamoeba isolates were found to be resistant to osmotic stress and heat shock, which supports their pathogenic potential. However, these findings highlight the need for improved disinfection protocols for sterile equipment.

Artificial intelligence systems (AI, specifically data-driven systems that generate predictions, classifications or recommendations based on statistical models and/or machine learning techniques) are increasingly used in service design and delivery, embedded in service processes, influencing how decisions are prepared, prioritised and executed. As their role becomes more prominent, it becomes relevant to clarify how responsibility is structured when humans and AI (or AI-enabled) systems both contribute to service outcomes.

Early and accurate diagnosis of prostate cancer remains challenging due to its biological heterogeneity and the limitations of current clinical tools. Widely used diagnostic approaches, including serum prostate-specific antigen (PSA) testing and biopsy, are associated with low specificity, invasiveness, and limited assessment of tumour aggressiveness. These limitations have driven increasing interest in non-invasive molecular diagnostics that can provide biologically meaningful insights while reducing unnecessary clinical interventions This review aims to evaluate the biological basis, analytical methodologies, and clinical utility of urine-based gene expression profiling for prostate cancer detection and risk stratification. Tumour-associated transcriptional alterations, persistent androgen receptor signalling, epigenetic dysregulation, metabolic reprogramming, and extracellular vesicle-mediated RNA release enable the detection of clinically relevant gene expression signatures in urine. Key urinary biomarkers include mRNAs, long non-coding RNAs, fusion transcripts, and microRNAs, which are analyzed using platforms ranging from quantitative PCR-based assays to next-generation sequencing and multigene classifier approaches. Evidence from prospective and multicentre clinical studies suggests that urine-based transcriptomic assays may improve the detection of clinically significant disease, reduce unnecessary biopsies, and support risk stratification when compared with PSA and conventional clinical models. However, biological variability, technical complexity, and standardization challenges remain important considerations. This review highlights current limitations and future perspectives, including the integration of urinary transcriptomic profiling with multimodal diagnostic strategies and artificial intelligence-based decision-support systems to advance precision prostate cancer care.

Obstetric and gynecological ultrasound has become one of the most important first-line diagnostic modalities in contemporary clinical practice, playing a central role in antenatal care, prenatal screening and gynecological diagnostics. The aim of this review is to provide a comprehensive overview of the development of obstetric and gynecological ultrasound in Hungary, its current national guideline framework and its integration into international standards, with particular emphasis on technological innovations and quality assurance. The article analyzes the position and relevance of Hungarian recommendations in comparison with major international guidelines, including those of ISUOG, FMF, ESHRE and the IOTA consortium, and summarizes the main clinical indications of ultrasound in obstetrics and gynecology. Special attention is given to the role of 3D/4D imaging, Doppler techniques, structured reporting and artificial intelligence-based decision support systems, which contribute significantly to improved diagnostic accuracy and reproducibility. The strengths of the Hungarian system include wide accessibility, guideline-based practice and a license-based training and competency framework. Future challenges involve further development of education, enhancement of auditability and the integration of data-driven and artificial intelligence-supported solutions into routine clinical workflows. The responsible and standardized use of modern ultrasound technology remains essential for patient safety, quality assurance and evidence-based clinical decision-making. Orv Hetil. 2026; 167(16): 610-620.

This study envisages to emphasize the significance of utilizing Artificial Intelligence in the currently prevailing banking and financial businesses. This article seeks to dissipate and address the possible ventures confederated with financial institutions which avail these technologies, customers or users, and investors including the market cyclicity and comprehensive risk. It emphasizes on the role of AI applications in enhancing financial organizations’ contentious interests which further obliges or compels the collaboration and participation of policymakers and regulators. The pertinence and affirmation of Artificial Intelligence systems in precipitating corporate operations, risk management, and revenue growth is gaining propulsion worldwide. This article examines the inference of rapid use of Artificial intelligence (AI) in the financial sector. It also focuses on the benefits of such technology with respect to financial depth and coherence while accentuating the scrutiny regarding the expansion of digital divide between developed and developing countries. The research imparts to the conversation about the effect of AI by emanating and tabulating the threats it may initiate with respect to the integrity and stability to the financial depth while also looking into the policy issues and effective regulatory measures. AI and its application in the financial sector are constantly growing but the entire extent of its strengths and disadvantages remains unidentified. Despite being provided the potential for unanticipated consequences, there is still a need to consolidate prudential monitoring. While attempting to highlight the consequences regarding implementation of AI and to determine the benefits and risks which accompanies the use of AI, this research work aims to arrive to a conclusion with proffering and recommendations for the regulatory bodies and the policymakers in the guise of responses and suggestions to pursue stimulating innovation of AI in finance to safeguard financial investors and consumers.

The rapid evolution of digital technologies has transformed health, mental health, and social care, offering new modalities of digital care, assistance, and support through web-based platforms, mobile apps, extended reality, wearables, and artificial intelligence systems. Despite this proliferation, there is little consensus on what constitutes "high-quality" digital care. Challenges persist regarding data security, interoperability, accessibility, sustainability, and professional competence, whereas existing standards and regulations provide fragmented guidance. This study aimed to develop a contextualized, consensus-based quality assessment framework for digital care, assistance, and support in Flanders, Belgium. For this purpose, perspectives across technology, organizational processes, and professional competencies were integrated. The study used a multiphase design comprising (1) 10 expert interviews with Flemish government officials; (2) a narrative literature review of 303 peer-reviewed and gray literature sources; (3) a 3-round Delphi study with 50 experts across 5 domains (end users, facilitators, technology developers, deontology and ethics experts, and digital inclusion and media literacy experts); and (4) 4 complementary focus groups and 3 interviews with specialists in artificial intelligence, regulation, social work, mental health, and IT. The Delphi rounds gathered iterative feedback through open-ended elicitation, structured rating, and classification of quality criteria. Quantitative data were analyzed using descriptive statistics, whereas qualitative feedback was subjected to thematic analysis. A total of 50 experts participated in round 1, a total of 40 (80%) participated in round 2, and 27 (54%) participated in round 3. Round 1 generated 577 unique quality criteria, consolidated into 26 clusters organized under 3 pillars: technology, organization, and professional competencies. The relative importance across pillars was balanced (mean score 37.29, SD 12.38 for technology; 33.33, SD 10.39 for professional competencies; and 29.80, SD 10.45 for organizations). Accessibility, reliability, and safety ranked highest for the technology; vision, quality monitoring, and infrastructure ranked highest for organization; and support, digital competencies, and ethics ranked highest for professional competencies. The finalized framework included 112 criteria, of which 35 (31.3%) were designated as optional and 77 (68.8%) were designated as minimum requirements. Focus groups and interviews validated the framework's comprehensiveness and usability, emphasizing proportional implementation, user centrality, and alignment with European Union regulations. Stakeholders highlighted the need for tools, training, and governance mechanisms to ensure adoption and sustainability. This study produced a codeveloped, context-sensitive quality assessment framework that balances technological robustness, organizational readiness, and professional competence in digital care, assistance, and support. The framework can serve both as a quality safeguard and a developmental road map. Accompanying self-assessment and governance tools enhance practical applicability. Implementation success will depend on governmental support, resource allocation, and structured feedback loops. Future research should pilot the framework in real-world settings, assess its impact, and establish mechanisms for continuous updates to maintain relevance in a rapidly evolving digital landscape.

ABSTRACT This article introduces the concept of AI harmscapes to examine how artificial intelligence systems reproduce and amplify structural violence, exclusion, and inequality in the Global South, while revealing spaces for transformative justice responses. Drawing on critical criminology and science and technology studies, we develop a framework for understanding AI as a governance tool that enhances assemblages of control while generating new forms of social harm. Through case studies of predictive policing in Cape Town, digital identity systems in Kenya, and facial recognition in India, we analyse how AI-enhanced technologies entrench existing inequalities while being contested by affected communities and civil society organizations. We challenge criminology’s reliance on post-hoc accountability mechanisms and argue for anticipatory, community-centred governance approaches grounded in nodal and polycentric governance principles. By centring Global South experiences and resistance strategies, this analysis contributes to digital criminology while advancing transformative approaches to justice in an increasingly algorithmic world.