Counter with Evidence! A Multi-Agent Memory Efficient Reasoning Framework for Hate Category Informed Counterspeech Generation
2026-08-24 • Computation and Language
Computation and Language
AI summaryⓘ
The authors created a system called FIRE that breaks down hate speech into five types and matches each with a fitting counterspeech reply. They also built a new dataset, FactualCS, which includes detailed labels and evidence to help the system learn better responses. FIRE outperforms previous methods by making replies that are more accurate, less toxic, and better targeted to the type of hate speech. Human judges preferred FIRE's responses, showing its potential for practical use. This work highlights the importance of understanding different kinds of hate speech to respond effectively and safely.
counterspeechhate speechmisinformationdehumanizationstereotypesdatasetnatural language generationtoxicity reductionfactualitymulti-agent framework
Authors
Sujoy Nath, Aswini Kumar, Tanmoy Chakraborty
Abstract
Counterspeech effectively neutralizes the impact of online hate. Although prior work explores automated counterspeech generation, it largely emphasizes stylistic control while treating hate speech as homogeneous, overlooking that distinct forms of abuse require fundamentally different counterspeech strategies. To address this gap, we introduce FIRE (Factuality Informed Multi-Agent Reasoning Framework) that first decomposes hate speech into one of the five distinct categories (misinformation, stereotype, conspiracy, dehumanizing, non-factual), and then maps it to a targeted counterspeech style. To facilitate FIRE, we curate FactualCS, a novel dataset of $4,784$ instances that provides the annotations regarding hate categories, reasoning traces, and evidence mappings, which are critical elements for grounded generation that are missing in prior work. A comprehensive evaluation across $28$ baseline configurations demonstrates that FIRE significantly surpasses existing methods, despite using compact agents ($<$2B). FIRE achieves a $\sim$ $12 \%$ and $\sim$ $11 \%$ improvements in factual and category-specific accuracy respectively, while simultaneously reducing toxicity by $\sim$ $11 \%$ relative to the strongest baselines. Further human evaluation confirms that responses generated by FIRE are significantly preferred over the strongest baselines, underscoring its effectiveness for real-world deployment. These findings show that decomposing the underlying intent of hate speech is essential for generating safe, effective, and contextually precise counterspeech.