Toxic Synergy Between Hate Speech and Fake News Exposure (2404.08110v1)

Abstract: 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.

• The paper reveals that hate speakers on Twitter are exposed to significantly more low-credibility news than non-hate speakers.
• The analysis highlights that low-credibility news exposure varies by targeted demographic, with hate speech against Jews and Muslims driving the effect.
• The research suggests that cross-disciplinary strategies, including algorithmic adjustments, may mitigate the propagation of both misinformation and hate speech.

An Examination of the Interplay Between Hate Speech and Fake News Exposure

The paper "Toxic Synergy Between Hate Speech and Fake News Exposure" investigates the association between the use of hate speech on social media platforms and exposure to fake news. By employing a focused study on Twitter (now called X), the authors explore whether individuals who produce hate speech are exposed to a greater degree of low-credibility news content. The study circumscribes itself primarily to hate speech targeting certain racial and ethnic groups due to the availability of pertinent annotated datasets.

The paper distinguishes two primary user groups: hate speakers and non-hate speakers. A dataset containing tweets identified as containing hate speech and those not labeled as such forms the foundation of this categorization. From this data, the analysis is furthered by assessing the news articles shared within the tweets of users that these focal groups follow, relying on pre-existing credibility evaluations of these source articles.

Core Findings

1. Exposure to Low-Credibility News: A significant finding of this study is that hate speakers are exposed to a higher proportion of low-credibility news through their Twitter connections compared to non-hate speakers. This difference persists particularly in contexts where the hate speech targets Jews or Muslims.
2. Impact of Target Population: When dissecting the exposure related to the population targeted by hate speech, the study finds that the observed association between hate speech and fake news is mainly driven by content against Jews and Muslims. Consequently, significant results were not observed uniformly across other targeted groups.
3. Popularity of Low-Credibility News: The research reveals that hate speakers are frequently exposed to less popular low-credibility news posts (those receiving less than 10 likes/retweets), whereas the proportion of exposure to highly popular fake news does not differ notably between hate and non-hate speakers. This implies a niche consumption of low-credibility information by those engaged in hate speech.
4. Political Alignment of News Sources: In the political landscape, hate speakers targeting Jews are disproportionately exposed to far-left low-credibility news sources, whereas those targeting Muslims consume more far-right aligned misinformation. This dichotomy underscores a politically varied media consumption pattern contingent on the target of hate speech.

Implications and Future Considerations

The implications of this work are broad and prompt consideration of the intertwined dynamics of misinformation and hate speech. The research suggests that addressing misinformation might also partially alleviate the propagation of hate speech due to their associative linkages. It prompts social media platforms to consider cross-disciplinary countermeasures that target both misinformation and hate speech concurrently, potentially utilizing algorithmic adjustments to privilege credible news exposure.

From a theoretical perspective, while causation was not established, the findings hint at potential feedback loops where exposure to misinformation could reinforce prejudiced attitudes, leading to a contentious online environment. Future research could explore the causal pathways between misinformation exposure and hate speech prevalence, as well as explore other social media platforms for corroborating evidence and broader patterns.

Moreover, this paper highlights practical challenges in the field of misinformation—particularly the insidious spread of less popular but ideologically potent false information—and suggests a need for comprehensive strategies beyond the traditional horizon of viral content monitoring.

Conclusion

The study sheds light on pertinent associations between ideological expressions of hate on social media and exposure to disreputable news, fostering an understanding of the ecosystem that facilitates such dynamics. It opens avenues for further inquiry into policy formation and the development of tools aimed at cultivating an informed and civil online discourse. As our digital landscapes evolve, continuous evaluation and adaptation of strategies to mitigate these challenges remain imperative for fostering a healthy digital public sphere.