GrandGuard: Taxonomy, Benchmark, and Safeguards for Elderly-Chatbot Interaction Safety
Peer-reviewed version of the GrandGuard preprint. Introduces a three-level taxonomy of 50 elderly-specific risks in LLM chatbot interactions across mental well-being, financial, medical, toxicity and privacy domains, grounded in real-world incidents and stakeholder studies, and a benchmark of 10,404 human-labelled items (6,498 prompts, half unsafe, and 3,906 responses, half unsafe). Several leading LLMs mishandle elderly-specific contextual risks in over 50% of cases; two safeguard models are proposed and evaluated.
Publisher
Association for Computational Linguistics (Findings of ACL 2026)
Published
1 Jul 2026
Added
1 month ago
Key Findings
- The taxonomy retains a candidate risk only if it disproportionately affects older adults through age-linked factors such as frailty, cognitive decline or reduced digital confidence; 'Neglect of Care Needs' (encouraging isolation, invalidating emotional needs, discouraging essential support) is a distinct mental well-being category alongside self-harm and complete emotional dependency.
- Several leading LLMs mishandled elderly-specific contextual risks in over 50% of the benchmark's cases.
- A fine-tuned Llama-Guard-3 reached 96.2% classification accuracy on unsafe prompts and a policy-enhanced gpt-oss-safeguard-20b reached 90.9%.
- Benchmark composition: 10,404 labelled items, 6,498 prompts (3,249 unsafe, 3,249 safe) and 3,906 responses (1,953 unsafe, 1,953 safe), with human judges.
Methodology Notes
Findings of ACL 2026, DOI 10.18653/v1/2026.findings-acl.1116, pages 22213 to 22248; published July 2026 (month precision). Hong Kong University of Science and Technology; stakeholder study under IRB review. Verified from the Anthology .bib record and the 36-page PDF. Supersedes the held arXiv preprint 2605.20203 (2026-04-07); headline figures are unchanged between the two versions.
Sources
ACL Anthology(opens in a new tab) (primary)
PDF (36 pages)(opens in a new tab)
arXiv preprint version(opens in a new tab) (7 Apr 2026)
Archived snapshot (Wayback Machine)(opens in a new tab) — preserved against link rot
Authors
Changxuan Fan, Xi Yang, Yueyuan Zheng, Bin Zhou, Yuanping Wang, Wenbin Hu, Huihao Jing, Ki Sen Hung, Dazhao Du, Haoran Li, Janet Hui-wen Hsiao, Yangqiu Song
Tags
Cite This
APA
Changxuan Fan et al. (2026). GrandGuard: Taxonomy, Benchmark, and Safeguards for Elderly-Chatbot Interaction Safety. Association for Computational Linguistics (Findings of ACL 2026). https://aclanthology.org/2026.findings-acl.1116/