Large Language Models (LLMs) have raised increasing concerns about their misuse in generating hate speech. Among all the efforts to address this issue, hate speech detectors play a crucial role. However, the effectiveness of different detectors against LLM-generated hate speech remains largely unknown. In this paper, we propose HateBench, a framework for benchmarking hate speech detectors on LLM-generated hate speech. We first construct a hate speech dataset of 7,838 samples generated by six widely-used LLMs covering 34 identity groups, with meticulous annotations by three labelers. We then assess the effectiveness of eight representative hate speech detectors on the LLM-generated dataset. Our results show that while detectors are generally effective in identifying LLM-generated hate speech, their performance degrades with newer versions of LLMs. We also reveal the potential of LLM-driven hate campaigns, a new threat that LLMs bring to the field of hate speech detection. By leveraging advanced techniques like adversarial attacks and model stealing attacks, the adversary can intentionally evade the detector and automate hate campaigns online. The most potent adversarial attack achie
Online hate speech has been linked to a global rise in violence against minorities, including incidents such as mass shootings, lynchings, and ethnic cleansing. Societies grappling with this issue, particularly when hate speech targets specific groups based on religion, race, ethnicity, culture, nationality, or migration status, face the challenge of balancing freedom of expression with the need for effective content moderation on widely used online platforms. In response to this challenge, we introduce a comprehensive hate speech dataset covering five distinct topics in Turkish: refugees, the Israel-Palestine conflict, anti-Greek sentiment in Turkey, ethnic or religious communities (Alevis, Armenians, Arabs, Jews, and Kurds), and LGBTI+, alongside one topic in Arabic (refugees). In addition, we develop state-of-the-art BERT-based models to address multiple dimensions of hate speech analysis, including hate category classification, hate intensity prediction, target identification, and hate speech span detection, enabling a comprehensive understanding of hateful content in online discourse.
Online hate spreads rapidly, yet little is known about whether preventive and scalable strategies can curb it. We conducted the largest randomized controlled trial of hate speech prevention to date: a 20-week messaging campaign on X in Nigeria targeting ethnic hate. 73,136 users who had previously engaged with hate speech were randomly assigned to receive prosocial video messages from Nigerian celebrities. The campaign reduced hate content by 2.5% to 5.5% during treatment, with about 75% of the reduction persisting over the following four months. Reaching a larger share of a user's audience reduced amplification of that user's hate posts among both treated and untreated users, cutting hate reposts by over 50% for the most exposed accounts. Scalable messaging can limit online hate without removing content.
Optimization of offensive content moderation models for different types of hateful messages is typically achieved through continued pre-training or fine-tuning on new hate speech benchmarks. However, existing benchmarks mainly address explicit hate toward protected groups and often overlook implicit or indirect hate, such as demeaning comparisons, calls for exclusion or violence, and subtle discriminatory language that still causes harm. While explicit hate can often be captured through surface features, implicit hate requires deeper, full-model semantic processing. In this work, we question the need for repeated fine-tuning and analyze the role of HatePrototypes, class-level vector representations derived from language models optimized for hate speech detection and safety moderation. We find that these prototypes, built from as few as 50 examples per class, enable cross-task transfer between explicit and implicit hate, with interchangeable prototypes across benchmarks. Moreover, we show that parameter-free early exiting with prototypes is effective for both hate types. We release the code, prototype resources, and evaluation scripts to support future research on efficient and tran
Hate speech on social media threatens the mental and physical well-being of individuals and contributes to real-world violence. Resharing is an important driver behind the spread of hate speech on social media. Yet, little is known about who reshares hate speech and what their characteristics are. In this paper, we analyze the role of user characteristics in hate speech resharing across different types of hate speech (e.g., political hate). For this, we proceed as follows: First, we cluster hate speech posts using large language models to identify different types of hate speech. Then we model the effects of user attributes on users' probability to reshare hate speech using an explainable machine learning model. To do so, we apply debiasing to control for selection bias in our observational social media data and further control for the latent vulnerability of users to hate speech. We find that, all else equal, users with fewer followers, fewer friends, fewer posts, and older accounts share more hate speech. This shows that users with little social influence tend to share more hate speech. Further, we find substantial heterogeneity across different types of hate speech. For example,
Hate speech remains prevalent in human society and continues to evolve in its forms and expressions. Modern advancements in internet and online anonymity accelerate its rapid spread and complicate its detection. However, hate speech datasets exhibit diverse characteristics primarily because they are constructed from different sources and platforms, each reflecting different linguistic styles and social contexts. Despite this diversity, prior studies on hate speech detection often rely on fixed methodologies without adapting to data-specific features. We introduce RV-HATE, a detection framework designed to account for the dataset-specific characteristics of each hate speech dataset. RV-HATE consists of multiple specialized modules, where each module focuses on distinct linguistic or contextual features of hate speech. The framework employs reinforcement learning to optimize weights that determine the contribution of each module for a given dataset. A voting mechanism then aggregates the module outputs to produce the final decision. RV-HATE offers two primary advantages: (1)~it improves detection accuracy by tailoring the detection process to dataset-specific attributes, and (2)~it a
Local or national politics can trigger potentially dangerous hate in someone. But with a third of the world's population eligible to vote in elections in 2024 alone, we lack understanding of how individual-level hate multiplies up to hate behavior at the collective global scale. Here we show, based on the most recent U.S. election, that offline events are associated with a rapid adaptation of the global online hate universe that hardens (strengthens) both its network-of-networks structure and the 'flavors' of hate content that it collectively produces. Approximately 50 million potential voters in hate communities are drawn closer to each other and to the broad mainstream of approximately 2 billion others. It triggers new hate content at scale around immigration, ethnicity, and antisemitism that aligns with conspiracy theories about Jewish-led replacement before blending in hate around gender identity/sexual orientation, and religion. Telegram acts as a key hardening agent - yet is overlooked by U.S. Congressional hearings and new E.U. legislation. Because the hate universe has remained robust since 2020, anti-hate messaging surrounding not only upcoming elections but also other eve
Recent advances in text-to-image diffusion models have enabled the creation of a new form of digital art: optical illusions--visual tricks that create different perceptions of reality. However, adversaries may misuse such techniques to generate hateful illusions, which embed specific hate messages into harmless scenes and disseminate them across web communities. In this work, we take the first step toward investigating the risks of scalable hateful illusion generation and the potential for bypassing current content moderation models. Specifically, we generate 1,860 optical illusions using Stable Diffusion and ControlNet, conditioned on 62 hate messages. Of these, 1,571 are hateful illusions that successfully embed hate messages, either overtly or subtly, forming the Hateful Illusion dataset. Using this dataset, we evaluate the performance of six moderation classifiers and nine vision language models (VLMs) in identifying hateful illusions. Experimental results reveal significant vulnerabilities in existing moderation models: the detection accuracy falls below 0.245 for moderation classifiers and below 0.102 for VLMs. We further identify a critical limitation in their vision encoder
We introduce a novel multi-labeled scheme for joint annotation of hate and counter-hate speech in social media conversations, categorizing hate and counter-hate messages into thematic and rhetorical dimensions. The thematic categories outline different discursive aspects of each type of speech, while the rhetorical dimension captures how hate and counter messages are communicated, drawing on Aristotle's Logos, Ethos and Pathos. We annotate a sample of 92 conversations, consisting of 720 tweets, and conduct statistical analyses, incorporating public metrics, to explore patterns of interaction between the thematic and rhetorical dimensions within and between hate and counter-hate speech. Our findings provide insights into the spread of hate messages on social media, the strategies used to counter them, and their potential impact on online behavior.
Many online hate groups exist to disparage others based on race, gender identity, sex, or other characteristics. The accessibility of these communities allows users to join multiple types of hate groups (e.g., a racist community and a misogynistic community), raising the question of whether users who join additional types of hate communities could be further radicalized compared to users who stay in one type of hate group. However, little is known about the dynamics of joining multiple types of hate groups, nor the effect of these groups on peripatetic users. We develop a new method to classify hate subreddits and the identities they disparage, then apply it to understand better how users come to join different types of hate subreddits. The hate classification technique utilizes human-validated deep learning models to extract the protected identities attacked, if any, across 168 subreddits. We find distinct clusters of subreddits targeting various identities, such as racist subreddits, xenophobic subreddits, and transphobic subreddits. We show that when users become active in their first hate subreddit, they have a high likelihood of becoming active in additional hate subreddits of
Counterspeech has gained attention as a strategy to reduce hate speech on social media. Although previous studies suggest that counterspeech can reduce hate speech, little is known about its effects on participation in online hate communities. Relatedly, we lack an understanding about the degree of hostility in counterspeech. Hostile counterspeech may increase online conflict, potentially hardening the positions of hate adherents, and further eroding online environments. Here, we analyzed the effect of counterspeech on 16,513 newcomers across 104 hate subreddits (forums within Reddit.com). We devised an LLM-based counterspeech detection approach that outperforms specialized models trained on existing datasets, then examined the presence, and effects of, hostility. While counterspeech comments are less toxic than hate speech comments, they are almost twice as toxic as other discourse within hate subreddits. We then evaluated the effect of counterspeech on newcomer engagement in hate subreddits. We found that newcomers using hate speech who receive counterspeech are less likely to continue posting within these hate subreddits, rather than becoming galvanized. We speculate that, inste
Hate speech on social media is a pressing concern. Understanding the factors associated with hate speech may help mitigate it. Here we explore the association between hate speech and exposure to fake news by studying the correlation between exposure to news from low-credibility sources through following connections and the use of hate speech on Twitter. Using news source credibility labels and a dataset of posts with hate speech targeting various populations, we find that hate speakers are exposed to lower percentages of posts linking to credible news sources. When taking the target population into account, we find that this association is mainly driven by anti-semitic and anti-Muslim content. We also observe that hate speakers are more likely to be exposed to low-credibility news with low popularity. Finally, while hate speech is associated with low-credibility news from partisan sources, we find that those sources tend to skew to the political left for antisemitic content and to the political right for hate speech targeting Muslim and Latino populations. Our results suggest that mitigating fake news and hate speech may have synergistic effects.
Curbing online hate speech has become the need of the hour; however, a blanket ban on such activities is infeasible for several geopolitical and cultural reasons. To reduce the severity of the problem, in this paper, we introduce a novel task, hate speech normalization, that aims to weaken the intensity of hatred exhibited by an online post. The intention of hate speech normalization is not to support hate but instead to provide the users with a stepping stone towards non-hate while giving online platforms more time to monitor any improvement in the user's behavior. To this end, we manually curated a parallel corpus - hate texts and their normalized counterparts (a normalized text is less hateful and more benign). We introduce NACL, a simple yet efficient hate speech normalization model that operates in three stages - first, it measures the hate intensity of the original sample; second, it identifies the hate span(s) within it; and finally, it reduces hate intensity by paraphrasing the hate spans. We perform extensive experiments to measure the efficacy of NACL via three-way evaluation (intrinsic, extrinsic, and human-study). We observe that NACL outperforms six baselines - NACL yi
Hate speech is a harmful form of online expression, often manifesting as derogatory posts. It is a significant risk in digital environments. With the rise of Large Language Models (LLMs), there is concern about their potential to replicate hate speech patterns, given their training on vast amounts of unmoderated internet data. Understanding how LLMs respond to hate speech is crucial for their responsible deployment. However, the behaviour of LLMs towards hate speech has been limited compared. This paper investigates the reactions of seven state-of-the-art LLMs (LLaMA 2, Vicuna, LLaMA 3, Mistral, GPT-3.5, GPT-4, and Gemini Pro) to hate speech. Through qualitative analysis, we aim to reveal the spectrum of responses these models produce, highlighting their capacity to handle hate speech inputs. We also discuss strategies to mitigate hate speech generation by LLMs, particularly through fine-tuning and guideline guardrailing. Finally, we explore the models' responses to hate speech framed in politically correct language.
Since the proliferation of social media usage, hate speech has become a major crisis. Hateful content can spread quickly and create an environment of distress and hostility. Further, what can be considered hateful is contextual and varies with time. While online hate speech reduces the ability of already marginalised groups to participate in discussion freely, offline hate speech leads to hate crimes and violence against individuals and communities. The multifaceted nature of hate speech and its real-world impact have already piqued the interest of the data mining and machine learning communities. Despite our best efforts, hate speech remains an evasive issue for researchers and practitioners alike. This article presents methodological challenges that hinder building automated hate mitigation systems. These challenges inspired our work in the broader area of combating hateful content on the web. We discuss a series of our proposed solutions to limit the spread of hate speech on social media.
Hate speech detection has been extensively studied, yet existing methods often overlook a real-world complexity: training labels are biased, and interpretations of what is considered hate vary across individuals with different cultural backgrounds. We first analyze these challenges, including data sparsity, cultural entanglement, and ambiguous labeling. To address them, we propose a culture-aware framework that constructs individuals' hate subspaces. To alleviate data sparsity, we model combinations of cultural attributes. For cultural entanglement and ambiguous labels, we use label propagation to capture distinctive features of each combination. Finally, individual hate subspaces, which in turn can further enhance classification performance. Experiments show our method outperforms state-of-the-art by 1.05\% on average across all metrics.
Reliable automatic hate speech (HS) detection systems must adapt to the in-flow of diverse new data to curtail hate speech. However, hate speech detection systems commonly lack generalizability in identifying hate speech dissimilar to data used in training, impeding their robustness in real-world deployments. In this work, we propose a hate speech generalization framework that leverages emotion knowledge in a multitask architecture to improve the generalizability of hate speech detection in a cross-domain setting. We investigate emotion corpora with varying emotion categorical scopes to determine the best corpus scope for supplying emotion knowledge to foster generalized hate speech detection. We further assess the relationship between using pretrained Transformers models adapted for hate speech and its effect on our emotion-enriched hate speech generalization model. We perform extensive experiments on six publicly available datasets sourced from different online domains and show that our emotion-enriched HS detection generalization method demonstrates consistent generalization improvement in cross-domain evaluation, increasing generalization performance up to 18.1% and average cro
Following the outbreak of a global pandemic, online content is filled with hate speech. Donald Trump's ''Chinese Virus'' tweet shifted the blame for the spread of the Covid-19 virus to China and the Chinese people, which triggered a new round of anti-China hate both online and offline. This research intends to examine China-related hate speech on Twitter during the two years following the burst of the pandemic (2020 and 2021). Through Twitter's API, in total 2,172,333 tweets hashtagged #china posted during the time were collected. By employing multiple state-of-the-art pretrained language models for hate speech detection, we identify a wide range of hate of various types, resulting in an automatically labeled anti-China hate speech dataset. We identify a hateful rate in #china tweets of 2.5% in 2020 and 1.9% in 2021. This is well above the average rate of online hate speech on Twitter at 0.6% identified in Gao et al., 2017. We further analyzed the longitudinal development of #china tweets and those identified as hateful in 2020 and 2021 through visualizing the daily number and hate rate over the two years. Our keyword analysis of hate speech in #china tweets reveals the most freque
Research on hate speech has predominantly revolved around detection and interpretation from textual inputs, leaving verbal content largely unexplored. While there has been limited exploration into hate speech detection within verbal acoustic speech inputs, the aspect of interpretability has been overlooked. Therefore, we introduce a new task of explainable audio hate speech detection. Specifically, we aim to identify the precise time intervals, referred to as audio frame-level rationales, which serve as evidence for hate speech classification. Towards this end, we propose two different approaches: cascading and End-to-End (E2E). The cascading approach initially converts audio to transcripts, identifies hate speech within these transcripts, and subsequently locates the corresponding audio time frames. Conversely, the E2E approach processes audio utterances directly, which allows it to pinpoint hate speech within specific time frames. Additionally, due to the lack of explainable audio hate speech datasets that include audio frame-level rationales, we curated a synthetic audio dataset to train our models. We further validated these models on actual human speech utterances and found th
The spread of COVID-19 has sparked racism and hate on social media targeted towards Asian communities. However, little is known about how racial hate spreads during a pandemic and the role of counterspeech in mitigating this spread. In this work, we study the evolution and spread of anti-Asian hate speech through the lens of Twitter. We create COVID-HATE, the largest dataset of anti-Asian hate and counterspeech spanning 14 months, containing over 206 million tweets, and a social network with over 127 million nodes. By creating a novel hand-labeled dataset of 3,355 tweets, we train a text classifier to identify hate and counterspeech tweets that achieves an average macro-F1 score of 0.832. Using this dataset, we conduct longitudinal analysis of tweets and users. Analysis of the social network reveals that hateful and counterspeech users interact and engage extensively with one another, instead of living in isolated polarized communities. We find that nodes were highly likely to become hateful after being exposed to hateful content. Notably, counterspeech messages may discourage users from turning hateful, potentially suggesting a solution to curb hate on web and social media platfor