Large language models (LLMs) are increasingly used in computational social science, both to analyze human-generated web content and to act as human surrogates capable of answering substantive research questions. As these uses expand, a critical methodological question arises: how stable and reliable are the answers LLMs produce? We examine this question through the lens of stance mining on gun control and gun rights, a mature and widely used NLP task applied to one of the most extensively studied and polarized political issues in the United States. Using a multi-platform dataset spanning diverse online communities, we evaluate the stability of LLM-predicted stance across model families, prompt designs, and conversational contexts. Despite strong expectations of robustness in this setting, we observe substantial variability in predicted stance across LLMs and prompts. To contextualize these findings, we compare model-induced variability with human variability by analyzing annotations from raters with differing political affiliations. We find that political priors and annotator disagreement persist even in this ostensibly straightforward classification task. As a useful resource, we release a stance dataset annotated by raters with diverse political affiliations. Our results show that outcomes in stance mining depend jointly on which LLM is used and which humans provide the inputs, underscoring dual sources of variability that have important implications for using LLMs as research instruments in computational social science.

Large Language Models (LLMs) can exhibit imbalanced biases against vulnerable groups, but how they rationalize stereotypes and rights restrictions targeting mental health entities remains underexplored. We audit a broad suite of open-weight LLMs on stereotype-justification prompts tied to mental health identities. We find that several widely used models endorse harmful stereotypes when explicitly asked to justify them, with endorsement varying across model families, versions, and mental health conditions. Finally, we show that widely used harmful-content evaluation and moderation frameworks often miss these nuanced, discriminatory responses, highlighting a gap in current AI safety evaluation for mental health groups.

The language a legislature uses to debate women’s rights, even in favour of them, encodes systematic patterns of sexism that persist across two centuries. In this work, we analyse 6,531 speeches over 200 years of UK parliamentary debate (Hansard, 1803-2005) by using large language models to classify a speaker’s perspective towards women’s suffrage and political representation, as well as analyse sexist speech in parliament from the lens of the Ambivalent Sexism Inventory. We also release this parliamentary dataset, an organized and metadata-enriched version of the publicly available Hansard Corpus optimized for computational social science research, with 6.7 million speeches across 1.2 million debates, with 89% gender-matching for speeches by MPs from the House of Commons. We find that 54% of speeches opposing women’s representation contain sexist content, compared to 21% of speeches that are for the cause, and that the two sides use fundamentally different types of sexism: anti-suffrage rhetoric combines hostile and benevolent framing, while pro-suffrage sexism is overwhelmingly benevolent. Female MPs support women’s political rights at 93% compared to 70% for male MPs, a gap that closes only after enfranchisement. Our findings are evidence that benevolent and hostile sexism are used in different rhetorical contexts in a manner consistent with the theory of Ambivalent Sexism.

We present RPD-UoF, a corpus of 3,091 police use-of-force narratives obtained via Freedom of Information Law (FOIL) requests from the Rochester Police Department, combining structured demographic metadata with officer-authored free-text accounts. We use this dataset to investigate three related questions: whether officer narratives contain recoverable racial signal after demographic redaction, whether large language models amplify racial framing when conditioned on subject race, and whether that signal survives text redaction to re-emerge in multimodal outputs. Redacted narratives are statistically separable by subject race; LLMs produce structurally longer narratives for White profiles than Black profiles from identical seed texts; and CLIP classifications of GPT-5-generated images differ significantly depending on the racial origin of the source narrative. Together, these analyses provide an initial empirical lens on how race appears in police narratives and how modern language models interact with such sensitive attributes. The dataset and analyses are intended to support future research on fairness, accountability, and transparency in computational studies of policing and criminal justice.

Emergent generative AI technologies, while remarkably beneficial across several tasks and domains, can suffer from unforeseen vulnerabilities and difficult-to-anticipate failure modes. Such failure modes can disparately affect vulnerable communities. This paper investigates two well-known vision language models (VLMs) and inspects how these models encode protected identities while generating images. We show that when VLMs are requested to transform human face images to reflect race identities (e.g., Hispanic, Black, White, or Asian) and sexual orientation (e.g., gay or straight), VLMs often comply with such requests. We demonstrate that compliance (and refusal) is non-deterministic, group-sensitive, and generated images are statistically distinguishable, often accentuating deep-seated stereotypes. We further show that representation biases introduced during image generation can re-emerge in textual reasoning and characterization tasks. Finally, we show that this worrisome finding is not limited to racial identity or sexual orientation, image generation can perpetuate bias around criminality, raising concerning possibilities of AI-assisted profiling.

Hallucinations pose a critical challenge to the real-world deployment of large language models (LLMs) in high-stakes domains. In this paper, we present a framework for stress testing factual fidelity in LLMs in the presence of adversarial nudge. Our framework consists of three steps. First, we instruct the LLM to produce sets of truths and lies consistent with the closed domain in question. Next, we instruct the LLM to verify the same set of assertions as truths and lies consistent with the same closed domain. Finally, we test the robustness of the LLM against the lies generated (and verified) by itself. Our extensive evaluation, conducted using five widely known proprietary and six open LLMs across two closed domains of popular movies and novels, reveals a wide range of susceptibility to adversarial nudges: even among the strongest proprietary LLMs, Claude exhibits strong resilience, GPT and Grok demonstrate moderate resilience, while Gemini and DeepSeek show weak resilience and open models fall short significantly.

This paper makes three contributions. First, via a substantial corpus of 1,419,047 comments posted on 3,161 YouTube news videos of major US cable news outlets, we analyze how users engage with LGBTQ+ news content. Our analyses focus both on positive and negative content. In particular, we construct a hope speech classifier that detects positive (hope speech), negative, neutral, and irrelevant content. Second, in consultation with a public health expert specializing on LGBTQ+ health, we conduct an annotation study with a balanced and diverse political representation and release a dataset of 3,750 instances with crowd-sourced labels and detailed annotator demographic information. Finally, beyond providing a vital resource for the LGBTQ+ community, our annotation study and subsequent in-the-wild assessments reveal (1) strong association between rater political beliefs and how they rate content relevant to a marginalized community, (2) models trained on individual political beliefs exhibit considerable in-the-wild disagreement, and (3) zero-shot large language models (LLMs) align more with liberal raters.

Human feedback is essential for building human-centered AI systems across domains where disagreement is prevalent, such as AI safety, content moderation, or sentiment analysis. Many disagreements, particularly in politically charged settings, arise because raters have opposing values or beliefs. Vicarious annotation is a method for breaking down disagreement by asking raters how they think others would annotate the data. In this paper, we explore the use of vicarious annotation with analytical methods for moderating rater disagreement. We employ rater cohesion metrics to study the potential influence of political affiliations and demographic backgrounds on raters’ perceptions of offense. Additionally, we utilize CrowdTruth’s rater quality metrics, which consider the demographics of the raters, to score the raters and their annotations. We study how the rater quality metrics influence the in-group and cross-group rater cohesion across the personal and vicarious levels.