Metraipolis › archive › dc_beb8dbe2b6fc

Whose Guardrail? User Identity Shapes the Effectiveness of Model Disclaimers in Mitigating LLM Political Persuasion — arXiv:2611.05347

Source
arXiv (en)
Original
https://arxiv.org/abs/2611.05347
License (as archived)
CC0-1.0
Archived
2026-11-04T21:50:41.929Z
SHA-256 of the archived text
59dc7156dc8e18fd16679326512bf8408ca3d01bc31b4261a86644fde1cc50e7
Stable URL of this page
https://metraipolis.com/documento/dc_beb8dbe2b6fc

How to cite

Whose Guardrail? User Identity Shapes the Effectiveness of Model Disclaimers in Mitigating LLM Political Persuasion — arXiv:2611.05347 — arXiv (en), https://arxiv.org/abs/2611.05347 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_beb8dbe2b6fc (Metraipolis archive dc_beb8dbe2b6fc, sha256 59dc7156dc8e18fd16679326512bf8408ca3d01bc31b4261a86644fde1cc50e7)
BibTeX
@misc{dcbeb8dbe2b6fc,
  title = {Whose Guardrail? User Identity Shapes the Effectiveness of Model Disclaimers in Mitigating LLM Political Persuasion — arXiv:2611.05347},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.05347},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_beb8dbe2b6fc; sha256 59dc7156dc8e18fd16679326512bf8408ca3d01bc31b4261a86644fde1cc50e7},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.05347. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Politically biased AI can shift users' opinions, yet it remains unclear whether model disclaimers protect against such persuasion. We report a preregistered experiment (N=1,586) examining how disclaimer design and user characteristics shape susceptibility to politically biased AI. We varied whether disclaimers identified political bias versus potential inaccuracies and whether they included an explanation. Participants completed two decision-making tasks with an LLM biased against their political views, with outcomes assessed immediately and one week later. Disclaimers identifying political bias and including an explanation reduced persuasion, but effects varied across contexts and did not persist. Individual differences were more consistent predictors: Republicans were more persuaded and reported higher trust than Democrats across tasks. Similarly, participants who used AI more frequently were also more persuaded, whereas those with greater self-reported AI knowledge were less persuaded. These findings suggest that generic disclaimers are not enough: effective safeguards must account for user heterogeneity and context.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.HC, nuovi lavori del giorno).
Categorie arXiv: Human-Computer Interaction (cs.HC).
Pagina del lavoro: https://arxiv.org/abs/2611.05347 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-04T21:38:34.506Z dalla pagina https://arxiv.org/list/cs.HC/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).