# A Multi-Stage Agentic Framework for Effective Counter-Narrative Generation and Refinement

arXiv: https://arxiv.org/abs/2609.14178 · HTML: https://mlsanity.com/paper/2609.14178/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Social simulation and multi-agent debate, one of 16 directions in LLM agents and tool use.
- Builds on 18 of its 82 references that are in the map.
- First posted on arXiv 12 Sep 2026.

## Placed in

- LLM agents and tool use / [Social simulation and multi-agent debate](https://mlsanity.com/agents/social-simulation-and-multi-agent-debate/)

## Builds on (most cited first, 18 of 18)

- [Generative Adversarial Networks](https://mlsanity.com/paper/2203.00667/) (2022)
- [Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities](https://mlsanity.com/paper/2507.06261/) (2025)
- [TweetEval: Unified Benchmark and Comparative Evaluation for Tweet Classification](https://mlsanity.com/paper/2010.12421/) (2020)
- [Fine-Grained Analysis of Propaganda in News Articles](https://mlsanity.com/paper/1910.02517/) (2019)
- [CheckEval: A reliable LLM-as-a-Judge framework for evaluating text generation using checklists](https://mlsanity.com/paper/2403.18771/) (2024)
- [Human-in-the-Loop for Data Collection: a Multi-Target Counter Narrative Dataset to Fight Online Hate Speech](https://mlsanity.com/paper/2107.08720/) (2021)
- [Generate, Prune, Select: A Pipeline for Counterspeech Generation against Online Hate Speech](https://mlsanity.com/paper/2106.01625/) (2021)
- [Persuade Me if You Can: A Framework for Evaluating Persuasion Effectiveness and Susceptibility Among Large Language Models](https://mlsanity.com/paper/2503.01829/) (2025)
- [LLM Can be a Dangerous Persuader: Empirical Study of Persuasion Safety in Large Language Models](https://mlsanity.com/paper/2504.10430/) (2025)
- [Argument Undermining: Counter-Argument Generation by Attacking Weak Premises](https://mlsanity.com/paper/2105.11752/) (2021)
- [The Moral Debater: A Study on the Computational Generation of Morally Framed Arguments](https://mlsanity.com/paper/2203.14563/) (2022)
- [The Levers of Political Persuasion with Conversational AI](https://mlsanity.com/paper/2507.13919/) (2025)
- [Russian propaganda on social media during the 2022 invasion of Ukraine](https://mlsanity.com/paper/2211.04154/) (2022)
- [ReZG: Retrieval-Augmented Zero-Shot Counter Narrative Generation for Hate Speech](https://mlsanity.com/paper/2310.05650/) (2023)
- [Can (A)I Change Your Mind?](https://mlsanity.com/paper/2503.01844/) (2025)
- [Northeastern Uni at Multilingual Counterspeech Generation: Enhancing Counter Speech Generation with LLM Alignment through Direct Preference Optimization](https://mlsanity.com/paper/2412.15453/) (2024)
- ["I understand your perspective": LLM Persuasion through the Lens of Communicative Action Theory](https://mlsanity.com/paper/2606.08076/) (2026)
- [Generating Grounded Responses to Counter Misinformation via Learning Efficient Fine-Grained Critiques](https://mlsanity.com/paper/2506.05924/) (2025)

## Abstract (arXiv)

The rapid diffusion of hate speech and misinformation on social networks challenges democratic societies, since direct suppression efforts may deepen polarization, fuel public distrusts, and strengthen extremist narratives. LLM-driven counter-narratives (CNs) offer a promising way to reduce those risks, yet their effectiveness depends on rhetorical and stylistic choices that remain poorly understood. We present a multi-stage agent-based framework for generating, refining, and evaluating CNs, applied to pro-Russian hate and misinformation narratives on the war with Ukraine and adaptable to other domains. A pilot experiment with human evaluators identifies effective technique style pairings, such as repetition with emotional framing enhancing persuasiveness. Building on these insights, we introduce a multi-agent refinement process that iteratively improves CNs for persuasiveness, emotional engagement, and shareability. After human validation confirmed improvement, an automated safety analysis shows that our refined CNs match or improve on expert-written counterspeech. A simulated experiment then shows that they reduce the perceived strength of pro-Russian narratives and consistently outperform a vanilla LLM baseline, highlighting a pathway toward scalable, narrative-specific interventions against hate speech and misinformation. Code and data accompanying this work are publicly available at https://github.com/carmelkron/inlg2026-counter-narratives.