The safety failures we are not instrumenting: a perspective on hidden safety-critical challenges in modern AI systems
2026-07-21 • Computers and Society
Computers and SocietyArtificial IntelligenceHuman-Computer Interaction
AI summaryⓘ
The authors explain that many AI safety problems are hidden and not obvious, unlike big or dramatic failures that get more attention. They say it's important to look at the whole system around AI, including how errors are noticed, handled, and fixed over time. They suggest a framework with five layers to understand these hidden risks, such as honesty about uncertainty, secure control, consistent safety over time, strong organizational oversight, and protecting the broader information environment. The authors also point out specific ways AI can fail quietly and offer ideas for better design and rules to improve overall trustworthiness.
AI safetysocio-technical systemsepistemic integritycontrol integritytemporal integrityorganizational integrityecosystem integrityreward hackingprompt injectionmodel collapse
Authors
Gjergji Kasneci, Enkelejda Kasneci
Abstract
Current AI safety discourse still focuses disproportionately on visible failures, including obvious harms, dramatic misuse, and hypothetical catastrophic scenarios. That focus is incomplete. In deployed systems, many of the most consequential failures are quieter: plausible rather than spectacular, distributed across components rather than localized in a single output, and normalized by workflows before they are recognized as hazards. We argue that a central safety challenge in modern AI systems is increasingly not only whether a model emits a harmful response, but whether the broader socio-technical system preserves the conditions under which errors remain visible, contestable, containable, and recoverable. We propose a five-layer framework for diagnosing these hidden risks: (1) epistemic integrity, concerning whether evidence and uncertainty are represented honestly enough to support calibrated reliance; (2) control integrity, concerning whether authority, permissions, and action boundaries remain robust under attack and optimization; (3) temporal integrity, concerning whether safety holds across sessions, memory updates, and deployment drift; (4) organizational integrity, concerning whether institutions retain the capacity to audit, assign responsibility, and intervene effectively; and (5) ecosystem integrity, concerning whether AI systems preserve rather than erode the information environment on which future oversight depends. Across these layers, we identify under-recognized risk patterns, including overreliance, uncertainty and legitimacy laundering in retrieval, prompt injection, reward hacking, memory poisoning, evaluation deception, fictional human oversight, synthetic evidence pollution, and model collapse. We conclude with design and governance recommendations and a research agenda for shifting AI safety from model-centric evaluation toward socio-technical reliability.