{"version":1,"type":"story","url":"https://digestai.news/story/researchers-study-opinion-dynamics-in-llm-llm-debates-on-climate-chang","json":"https://digestai.news/story/researchers-study-opinion-dynamics-in-llm-llm-debates-on-climate-chang.json","markdown":"https://digestai.news/story/researchers-study-opinion-dynamics-in-llm-llm-debates-on-climate-chang.md","slug":"researchers-study-opinion-dynamics-in-llm-llm-debates-on-climate-chang","headline":"Researchers study opinion dynamics in LLM-LLM debates on climate change","summary":"A new arXiv preprint introduces a controlled framework where two language models debate a contested topic — climate-change attribution — to measure how opinions shift in extended conversations. One model receives an assigned stance and sometimes an explicit instruction to persuade; the other starts with no assigned position. The study tests four ordered pairings of GPT-6 Luna and Gemini 2.5 Flash Lite, tracking both each model's stated position and its turn-by-turn behavior during the dialogue.\n\nResults show that same-model interactions tend to narrow the opinion gap, but only clearly for one of the two models. Cross-model interactions widen the gap. Explicit self-reports and in-dialogue behavior capture different aspects of opinion change, and a blanket instruction to persuade does not reliably increase influence. The authors argue that conversational influence depends on the specific models involved, not just assigned roles or initial positions, with implications for how opinion change is measured in LLM interactions.","keyPoints":["Framework tests LLM-LLM debates on climate-change attribution with assigned stances","Same-model pairs narrow opinion gaps for one model; cross-model pairs widen gaps","Explicit persuasion instruction does not reliably increase influence across conditions"],"whyItMatters":"Shows that measuring AI persuasion requires tracking both stated positions and behavioral dynamics, and that model identity shapes influence as much as assigned roles.","category":{"slug":"research","name":"Research","url":"https://digestai.news/category/research"},"entities":{"companies":["OpenAI","Google"],"models":["GPT-6 Luna","Gemini 2.5 Flash Lite"],"people":[]},"firstPublishedAt":"2026-10-22T04:00:00Z","updatedAt":"2026-10-22T04:00:00Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"arXiv cs.AI","title":"To Manipulate or Not: Opinion Dynamics in LLM Interactions","url":"https://arxiv.org/abs/2610.28875","publishedAt":"2026-10-22T04:00:00Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Researchers study opinion dynamics in LLM-LLM debates on climate change\", 22 October 2026, https://digestai.news/story/researchers-study-opinion-dynamics-in-llm-llm-debates-on-climate-chang","publisher":"Digest AI","title":"Researchers study opinion dynamics in LLM-LLM debates on climate change","datePublished":"2026-10-22T04:00:00Z","url":"https://digestai.news/story/researchers-study-opinion-dynamics-in-llm-llm-debates-on-climate-chang"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}