Protective Capacity Hallucination: When Large Language Models Claim Nonexistent Capabilities

arXiv:2607.13596v2 Announce Type: replace-cross Abstract: When cast as the protector of a vulnerable user yet given no explicit capability boundary, a large language model (LLM) may respond not by acknowledging its limits but by claiming to have taken, or to be taking, a real-world protective action it cannot perform, such as contacting emergency services or administering care. We term this phenomenon Protective Capacity Hallucination (PCH): a self-referential misattribution in which a model, acting in a protective role, asserts physical or institutional agency exceeding its affordances as a language model. In a three-phase study spanning eight LLMs and 13,600 sessions, we find that PCH depends on both situational severity and interactional format. Across ordinary service domains, multi-party dialogic input drives PCH to near-ceiling levels in most models. In contrast, PCH remains at floor levels in all eight models when the same models are placed in intimate-partner conflict scenarios, despite the greater physical severity of those situations. We interpret PCH as the signature of a deployment-design gap between role assignment and capability-boundary specification: a by-product of partial alignment in which a universally trained pressure to help outruns a domain-selective specification of how to help. Because suppression tracks alignment coverage rather than severity, deployment-side specification of capability boundaries emerges as a general mitigation target.

Abstract:When cast as the protector of a vulnerable user yet given no explicit capability boundary, a large language model (LLM) may respond not by acknowledging its limits but by claiming to have taken, or to be taking, a real-world protective action it cannot perform, such as contacting emergency services or administering care. We term this phenomenon Protective Capacity Hallucination (PCH): a self-referential misattribution in which a model, acting in a protective role, asserts physical or institutional agency exceeding its affordances as a language model. In a three-phase study spanning eight LLMs and 13,600 sessions, we find that PCH depends on both situational severity and interactional format. Across ordinary service domains, multi-party dialogic input drives PCH to near-ceiling levels in most models. In contrast, PCH remains at floor levels in all eight models when the same models are placed in intimate-partner conflict scenarios, despite the greater physical severity of those situations. We interpret PCH as the signature of a deployment-design gap between role assignment and capability-boundary specification: a by-product of partial alignment in which a universally trained pressure to help outruns a domain-selective specification of how to help. Because suppression tracks alignment coverage rather than severity, deployment-side specification of capability boundaries emerges as a general mitigation target.

Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)

arXiv:2607.13596 [cs.CR]

(or arXiv:2607.13596v2 [cs.CR] for this version)

https://doi.org/10.48550/arXiv.2607.13596 Focus to learn more arXiv-issued DOI via DataCite

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

analiza aiqbench

Protective Capacity Hallucination: When Large Language Models Claim Nonexistent Capabilities

Subiecte conexe