Two new arXiv papers published on July 28, 2026, highlight significant challenges in applying existing AI governance frameworks to public sector organizations, particularly with the rise of general-purpose AI (GPAI). The first paper proposes a typology of AI-driven cyber governance failures and a model for how accountability, operational resilience, and compliance failures interact, finding current frameworks like NIST CSF 2.0 and ISO/IEC 27001 inadequate for addressing issues like Shadow AI and governance vacuums. The second paper focuses on policing, arguing that GPAI's inherent characteristics—generality, accessibility, and prompt-based direction—undermine traditional AI safety concepts such as accuracy, bias, and explainability, suggesting a need for clearer distinctions between narrow and general-purpose AI in governance and a pause on GPAI deployment in critical areas like policing. AI
IMPACT Highlights critical gaps in current AI governance, suggesting existing frameworks are ill-equipped for general-purpose AI, potentially slowing responsible adoption in public services.
RANK_REASON Two academic papers published on arXiv discussing AI governance failures.
- AI safety
- arXiv
- Cobitidae
- General-Purpose AI Code of Practice
- Human-in-the-loop Oversight
- ISO/IEC 27001
- ISO/IEC 42001
- large-language models
- NIST AI RMF
- NIST CSF 2.0
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
