All traces

Generative AI integration in higher education threatens academic integrity and equity by undermining authenticity of student work and widening inequalities.

Mainstream implementation of AI in healthcare is hindered by data security issues and budget and resource constraints.

AI for extreme climate events is limited by noisy, heterogeneous, small sample sizes with limited annotations, challenges integrating real-time information, and lack of understandable models needed for stakeholder trust and regulatory compliance.

The same AI-driven neuroadaptive learning systems raise implementation problems for K-12 and adult learners, including data privacy and data security risks, ethical concerns and algorithmic bias, scalability issues, and accessibility disparities.

AI adoption in nursing faces barriers including data privacy risks, algorithmic bias, lack of transparency and accountability, resistance to adoption, and disparities in access to AI technologies and standardized education.

The original 2019 PROBAST tool had become outdated given rapid progress in prediction modelling methodology and AI including machine learning.

Current automated and human moderation systems on Roblox fail to prevent young users' exposure to inappropriate content, cyberbullying, and predatory behavior.

ML-enhanced point-of-care testing faces regulatory hurdles, reliability questions, and privacy concerns that limit widespread clinical adoption.

Deploying IoT-enabled smart sensors with AI for precision agriculture is limited by high initial investment costs, complexities in data management, requirements for technical expertise, data security and privacy concerns, and connectivity issues in remote agricultural areas.

Medical LLMs still face numerous challenges in practical applications, including hallucination, limited interpretability, and ethical concerns that hinder widespread use.

Digital mental health tools are hampered by engagement challenges, industry setbacks, methodological critiques, and gaps in evidence and scaling that limit real-world applicability.

LLM summarisation of consultations showed a 1.47% hallucination rate and 3.45% omission rate, creating fidelity gaps that could compromise patient safety.

Fraud-detection models trained for specific scam types often fail to generalize to new fraud types and lose effectiveness when trained on outdated data, with inconsistent performance reporting.

AI-driven wearables face technical, ethical and regulatory hurdles including data interoperability, privacy concerns, algorithmic bias, scalability and security that limit widespread clinical adoption.

Current GDPR and AI Act definitions of automated decision-making fail to capture real-life applications where automated systems assist rather than replace human decision-makers in migration and asylum.