arXiv subjects

Gabriele Lenzini

Publications and source records attributed to Gabriele Lenzini.

At least 19 recordsLinked to original sources

arXiv2026

InferD: A Honeypot Framework for AI Inference Infrastructure

Large language model serving stacks, commercial inference APIs, vector databases, and agent orchestration gateways are now ex- posed on the public internet and indexed by commodity scanners within hours of deployment. The threat characteristics of this sur- face are only beginning to be measured, and the honeypot frame- works that could measure it were designed for conventional net- work services rather than for AI infrastructure. In fact, emulating an AI service is a distinct deception problem: the observable behavior of the service, including model inventories, token formats, tool-call protocols, and authorization flows, must be reproduced faithfully, while the underlying computation, model inference and tool ex- ecution, must never occur. We present InferD, a low-interaction honeypot framework organized around a single principle we call deception without inference: every request an adversary can observe is answered from static response pools, and nothing an adversary submits is ever executed. InferD decomposes the emulation into four deception planes, transport, service protocol, vulnerability surface, and agentic session, adds a provenance plane that links a planted credential to evidence of downstream model-mediated processing.

cs.CR↗

arXiv2026

Needs Your Help: Understanding Platform-Directed Rating Participation in Community Notes on X

Community-based fact-checking is promising in countering misinformation, yet its scalability is constrained by slow rating accumulation. To address this challenge, platforms such as X implement platform-directed rating mechanisms, specifically through ``Needs Your Help'' algorithmic prompts, to target unresolved notes. Using a dataset of over 220 million rating contributions -- including 1.9 million platform-directed ratings -- on X, we examine contributors' response to note prompts, notes' resolution, and raters' spillovers. We found (i) at note level, population-sampled ratings concentrate on recent notes with certain helpfulness and high disagreement. Once sampled, population-sampled rating was associated with faster and more transitions to resolved statuses. (ii) At rater level, following raters' first observed population-sampled rating, raters exhibit significant yet modest increases in daily ratings, rating pace and tag usage, while other behaviors show no change. These highlight the promise of algorithmic nudges to guide volunteer attention toward contested content, accelerating consensus while sustaining rater engagement.

cs.HC↗

arXiv2026

Fact-Checking Comments Precede Community Notes to Misleading Posts on X

Community-based fact-checking systems, such as X's Community Notes program, offer a potentially scalable approach against online misinformation diffusion. However, their efficacy is substantially undermined by the delay of fact-check delivery. To address this challenge, we investigated the promise of organic user comments as rapid corrective signals by analyzing a large-scale dataset of over 2.2 million comments directed to 1,841 community fact-checked misleading posts. We developed a high-performance language model pipeline to identify ``fact-checking (FC) comments'', i.e., comments that correct source posts with reasoning or evidence (91% accuracy). Using this pipeline, we find that 99.4% of misleading posts receive their first FC comments before the official community note is created. Notably, the median time to the initial FC comment is only 0.1 hours, while the creation of community notes has a median latency of 9.7 hours since the publication of misleading posts. Additionally, compared to random posts, those misleading posts with displayed notes have significantly more FC comments. This suggests that FC comments are a unique pattern for misleading posts. We further identified characteristics of FC comments' intensity and speed: (i) FC comments' volume is significantly associated with content richness and misinformation type, favoring multimedia content and missing-context scenarios; (ii) the speed of FC comments remains robust across diverse topics and emotional intensities. These results suggest that platforms could use FC comments as early-warning signals, and potential resource for complementing community notes or automated synthesis.

cs.HC↗

arXiv2026

Community corrections have divergent downstream effects across corrected accounts

Community-based fact-checking can reduce the spread of annotated misleading posts, but whether it produces lasting behavioral change among corrected authors remains unclear. Here, we conduct a large-scale quasi-experimental study of the Community Notes system on X (formerly Twitter), tracking four weeks of activity before and after note display for 19,854 accounts and 57,935 corrections (noted posts and matched controls), covering 11,909,591 original posts. Difference-in-Differences estimates show that note display is followed by an average 2.9% increase in corrected accounts' original-post activity. This aggregate conceals two divergent trajectories. Accounts corrected only once reduce their activity by 2.4% and subsequently publish less toxic and less misleading content. Repeatedly corrected accounts, which constitute 28.6% of corrected accounts but produce 73.4% of fact-checked posts, instead increase their activity by 4.4% after their first correction and show no detectable response to later ones. They exhibit no comparable content improvement, and instead publish more highly misleading posts, cite lower-quality domains, and post more political content. Community notes can thus constrain individual misleading posts without durably improving the behavior of the accounts most responsible for them, indicating that correcting content and changing its producers are distinct objectives for platform design.

cs.SI↗

arXiv2026

Hiding Liker Identity Did Not Increase Engagement With Reputationally Risky Content on X (Formerly Twitter)

In June 2024, X (formerly Twitter) made likes from public to private, offering a rare, platform-level opportunity to study how the visibility of engagement signals affects users' behavior. Here, we investigate whether hiding liker identities increases the number of likes received by high-reputational-risk content, content for which public endorsement may carry high social or reputational costs due to its topic (e.g., politics) or the account context in which it appears (e.g., partisan accounts). To this end, we conduct two complementary studies: 1) a Difference-in-Differences analysis of 153,704 posts that are created by 1045 accounts and have received over 324 million likes on X (formerly Twitter) before and after the policy change; 2) a within-subject survey experiment with 203 X (formerly Twitter) users on participants' self-reported willingness to like different kinds of content. We find no detectable platform-level increase in likes for high-reputational-risk content (Study 1). Additionally, while participants in the survey experiment, particularly those with higher education and income, report modest increases in willingness to like high-reputational-risk content under private versus public visibility, these increases do not lead to significant changes in the group-level average likelihood of liking posts (Study 2). Taken together, our results suggest that hiding liker identity produces a limited behavioral response at the platform level, which may be caused by a gap between user intention and behavior.

cs.SI↗

arXiv2026

Poison to Detect: Detection of Targeted Overfitting in Federated Learning

Federated Learning (FL) enables collaborative model training among clients without centralising data, making it a widely adopted privacy-enhancing technology (PET). Despite its privacy benefits, FL remains vulnerable to orchestrator-driven privacy attacks. In this paper, we study an underexplored threat in which a dishonest orchestrator intentionally manipulates the aggregation process to induce targeted overfitting in local models of specific clients. Although prior work focuses on reducing information leakage during training, we emphasise early client-side detection of targeted overfitting, allowing clients to disengage before significant harm occurs. To this end, we propose three detection techniques -- label flipping, backdoor trigger injection, and model fingerprinting -- which enable clients to verify the integrity of the global aggregation. We evaluated our methods across multiple datasets and attack scenarios. In single-client attacks, all three methods detect orchestrator-induced overfitting within 1-2 training rounds with F1 scores up to 0.7. Scalability experiments further show that detection effectiveness is influenced by cohort composition and method parameters. These results demonstrate that client-side integrity testing can provide early, effective, and scalable detection, supporting safer deployment of FL systems.

cs.CR↗

arXiv2026

Request a Note: How the Request Function Shapes X's Community Notes System

X's Community Notes is a crowdsourced fact-checking system. To improve its scalability, X introduced ``Request Community Note'' feature, enabling users to solicit fact-checks from contributors on specific posts. Yet, its implications for the system -- what gets checked, by whom, and with what quality -- remain unclear. Using 98,685 requested posts and their associated notes, we evaluate how requests shape the Community Notes system. We find that requested posts with higher GPT-estimated misleadingness and from authors with greater misinformation exposure are more likely to receive notes. Conversely, requested political posts (vs. non-political) are less likely to receive notes. We also observe partisan asymmetries: posts from Republicans are more likely to receive notes than those from Democrats. Although only 12% of requested posts receive request-fostered notes from top contributors, these notes are rated as more helpful and less polarized than others, partly reflecting top contributors' selective fact-checking of misleading posts. Our findings highlight both the limitations and promise of requests for scaling high-quality community-based fact-checking.

cs.SI↗

arXiv2026

Can Theory-Informed Message Framing Drive Honest and Motivated Performance with Better Assessment Experiences in a Remote Assessment?

Remote unproctored assessments increasingly use messaging interventions to reduce cheating, but existing approaches lack theoretical grounding, focus narrowly on cheating suppression while overlooking performance and experience, and treat cheating as binary rather than continuous. This study examines whether messages based on 15 psychological concepts from self-determination, cognitive dissonance, social norms, and self-efficacy theories can reduce cheating while preserving performance and experience. Through an expert workshop (N=5), we developed 45 theory-informed messages and tested them with online participants (N=1232) who completed an incentivized anagram task. Participants were classified as non-cheaters (0% items cheated), partial-cheaters (1-99% cheated), or full-cheaters (100% cheated). Results show that concept-based messages reduced full-cheating occurrence by 42% (33% to 19%), increased non-cheating by 19% (53% to 63%), with no negative effects on performance or experience across integrity groups. Surprisingly, messages grounded in different theoretical concepts produced virtually identical effects. Analyses of self-rated psychological mechanisms revealed that messages influenced multiple mechanisms simultaneously rather than their intended targets, though these mechanisms predicted behavior, performance, and experience. These findings show that causal pathways are more complex than current theories predict. Practically, integrity interventions using supportive motivation rather than rule enforcement can reduce cheating without impairing performance or experience.

cs.HC↗

arXiv2026

Consensus Stability of Community Notes on X

Community-based fact-checking systems, such as Community Notes on X (formerly Twitter), aim to mitigate online misinformation by surfacing annotations judged helpful by contributors with diverse viewpoints. While prior work has shown that the platform's bridging-based algorithm effectively selects helpful notes at the time of display, little is known about how evaluations change after notes become visible. Using a large-scale dataset of 437,396 community notes and 35 million ratings from over 580,000 contributors, we examine the stability of helpful notes and the rating dynamics that follow their initial display. We find that 30.2% of displayed notes later lose their helpful status and disappear. Using interrupted time series models, we further show that note display triggers a sharp increase in rating volume and a significant shift in rating leaning, but these effects differ across rater groups. Contributors with viewpoints similar to note authors tend to increase supportive ratings, while dissimilar contributors increase negative ratings, producing systematic post-display polarization. Counterfactual analyses suggest that this post-display polarization, particularly from dissimilar raters, plays a substantial role in note disappearance. These findings highlight the vulnerability of consensus-based fact-checking systems to polarized rating behavior and suggest pathways for improving their resilience.

cs.SI↗

arXiv2025

From News Source Sharers to Post Viewers: How Topic Diversity and Conspiracy Theories Shape Engagement With Misinformation During a Health Crisis

Online engagement with misinformation threatens societal well-being, particularly during health crises when susceptibility to misinformation is heightened in a multi-topic context. Here, we focus on the COVID-19 pandemic and address a critical gap in understanding engagement with multi-topic misinformation on social media at two user levels: news source sharers (who post news items) and post viewers (who engage with news posts). To this end, we analyze 7273 fact-checked source news items and their associated posts on X through the lens of topic diversity and conspiracy theories. We find that false news, especially those containing conspiracy theories, exhibits higher topic diversity than true news. At news source sharer level, false news has a longer lifetime and receives more posts on X than true news, with conspiracy theories further extending its longevity. However, topic diversity does not significantly influence news source sharers' engagement. At post viewer level, contrary to news source sharer level, posts characterized by heightened topic diversity receive more reposts, likes, and replies. Notably, post viewers tend to engage more with misinformation containing conspiracy narratives: false news posts that contain conspiracy theories, on average, receive 40.8% more reposts, 45.2% more likes, and 44.1% more replies compared to those without conspiracy theories. Our findings suggest that news source sharers and post viewers exhibit distinct engagement patterns on X, offering valuable insights into refining misinformation interventions at these two user levels.

cs.SI↗

arXiv2025

Identifying Ethical Challenges in XR Implementations in the Industrial Domain: A Case of Off-Highway Machinery

Although extended reality(XR)-using technologies have started to be discussed in the industrial setting, it is becoming important to understand how to implement them ethically and privacy-preservingly. In our paper, we summarise our experience of developing XR implementations for the off-highway machinery domain by pointing to the main challenges we identified during the work. We believe that our findings can be a starting point for further discussion and future research regarding privacy and ethical challenges in industrial applications of XR.

cs.HC↗

arXiv2025

More Than Just Warnings:Exploring the Ways of Communicating Credibility Assessment on Social Media

Reducing the spread of misinformation is challenging. AI-based fact verification systems offer a promising solution by addressing the high costs and slow pace of traditional fact-checking. However, the problem of how to effectively communicate the results to users remains unsolved. Warning labels may seem an easy solution, but they fail to account for fuzzy misinformation that is not entirely fake. Additionally, users' limited attention spans and social media information should be taken into account while designing the presentation. The online experiment (n = 537) investigates the impact of sources and granularity on users' perception of information veracity and the system's usefulness and trustworthiness. Findings show that fine-grained indicators enhance nuanced opinions, information awareness, and the intention to use fact-checking systems. Source differences had minimal impact on opinions and perceptions, except for informativeness. Qualitative findings suggest the proposed indicators promote critical thinking. We discuss implications for designing concise, user-friendly AI fact-checking feedback.

cs.HC↗

arXiv2025

Community Fact-Checks Trigger Moral Outrage in Replies to Misleading Posts on Social Media

Displaying community fact-checks is a promising approach to reduce engagement with misinformation on social media. However, how users respond to misleading content emotionally after community fact-checks are displayed on posts is unclear. Here, we employ quasi-experimental methods to causally analyze changes in sentiments and (moral) emotions in replies to misleading posts following the display of community fact-checks. Our evaluation is based on a large-scale panel dataset comprising N=2,225,260 replies across 1841 source posts from X's Community Notes platform. We find that informing users about falsehoods through community fact-checks significantly increases negativity (by 7.3%), anger (by 13.2%), disgust (by 4.7%), and moral outrage (by 16.0%) in the corresponding replies. These results indicate that users perceive spreading misinformation as a violation of social norms and that those who spread misinformation should expect negative reactions once their content is debunked. We derive important implications for the design of community-based fact-checking systems.

cs.SI↗

arXiv2024

Changes in Sentiments and User Engagement for 2024 U.S. Presidential Candidates After Biden's Withdrawal: An Analysis of TikTok Videos

The 2024 U.S. presidential election has sparked widespread online discussions about the presidential candidates. Joe Biden's withdrawal from the race and Kamala Harris's subsequent entry as the Democratic candidate likely alter the dynamics of these online discussions; yet, this hypothesis requires evidence. Here, we study how sentiments and user engagement in social media posts mentioning presidential candidates change after Biden's withdrawal. Our analysis is based on N=680,609 TikTok videos that have accumulated over 4 billion views, with more than 23 million comments, 31 million shares, and 335 million likes from November 1, 2023, to October 6, 2024. We find that: (i) Before Biden's withdrawal, video posts mentioning the Republican candidate (Donald Trump) have higher positive sentiment and lower negative sentiment compared to those mentioning the Democratic candidate (Joe Biden). (ii) Following Biden's withdrawal, positive sentiment in video posts mentioning the Democratic candidate (Kamala Harris) increases by 46.8%, while negative sentiment decreases by 52.0%. (iii) Regarding user engagement, before Biden's withdrawal, video posts mentioning the Democratic candidate have 64.9% higher odds of being shared and 39.5% higher odds of receiving likes compared to posts mentioning the Republican candidate, with similar odds of receiving comments. (iv) After Biden's withdrawal, the odds of being shared increase by 53.3%, and the odds of receiving likes increase by 77.4% in both video posts mentioning the Democratic candidate and video posts mentioning the Republican candidate. Our findings offer insights into how sentiments and user engagement in online posts about the 2024 U.S. presidential candidates shift following Biden's dropping out from the presidential race.

cs.SI↗

arXiv2024

Is Fact-Checking Politically Neutral? Asymmetries in How U.S. Fact-Checking Organizations Pick Up False Statements Mentioning Political Elites

Political elites play an important role in the proliferation of online misinformation. However, an understanding of how fact-checking platforms pick up politicized misinformation for fact-checking is still in its infancy. Here, we conduct an empirical analysis of mentions of U.S. political elites within fact-checked statements. For this purpose, we collect a comprehensive dataset consisting of 35,014 true and false statements that have been fact-checked by two major fact-checking organizations (Snopes, PolitiFact) in the U.S. between 2008 and 2023, i.e., within an observation period of 15 years. Subsequently, we perform content analysis and explanatory regression modeling to analyze how veracity is linked to mentions of U.S. political elites in fact-checked statements. Our analysis yields the following main findings: (i) Fact-checked false statements are, on average, 20% more likely to mention political elites than true fact-checked statements. (ii) There is a partisan asymmetry such that fact-checked false statements are 88.1% more likely to mention Democrats, but 26.5% less likely to mention Republicans, compared to fact-checked true statements. (iii) Mentions of political elites in fact-checked false statements reach the highest level during the months preceding elections. (iv) Fact-checked false statements that mention political elites carry stronger other-condemning emotions and are more likely to be pro-Republican, compared to fact-checked true statements. In sum, our study offers new insights into understanding mentions of political elites in false statements on U.S. fact-checking platforms, and bridges important findings at the intersection between misinformation and politicization.

cs.SI↗

arXiv2024

Community-based fact-checking reduces the spread of misleading posts on social media

Community-based fact-checking is a promising approach to verify social media content and correct misleading posts at scale. Yet, causal evidence regarding its effectiveness in reducing the spread of misinformation on social media is missing. Here, we performed a large-scale empirical study to analyze whether community notes reduce the spread of misleading posts on X. Using a Difference-in-Differences design and repost time series data for N=237,677 (community fact-checked) cascades that had been reposted more than 431 million times, we found that exposing users to community notes reduced the spread of misleading posts by, on average, 62.0%. Furthermore, community notes increased the odds that users delete their misleading posts by 103.4%. However, our findings also suggest that community notes might be too slow to intervene in the early (and most viral) stage of the diffusion. Our work offers important implications to enhance the effectiveness of community-based fact-checking approaches on social media.

cs.SI↗

arXiv2024

Did the Roll-Out of Community Notes Reduce Engagement With Misinformation on X/Twitter?

Developing interventions that successfully reduce engagement with misinformation on social media is challenging. One intervention that has recently gained great attention is X/Twitter's Community Notes (previously known as "Birdwatch"). Community Notes is a crowdsourced fact-checking approach that allows users to write textual notes to inform others about potentially misleading posts on X/Twitter. Yet, empirical evidence regarding its effectiveness in reducing engagement with misinformation on social media is missing. In this paper, we perform a large-scale empirical study to analyze whether the introduction of the Community Notes feature and its roll-out to users in the U.S. and around the world have reduced engagement with misinformation on X/Twitter in terms of retweet volume and likes. We employ Difference-in-Differences (DiD) models and Regression Discontinuity Design (RDD) to analyze a comprehensive dataset consisting of all fact-checking notes and corresponding source tweets since the launch of Community Notes in early 2021. Although we observe a significant increase in the volume of fact-checks carried out via Community Notes, particularly for tweets from verified users with many followers, we find no evidence that the introduction of Community Notes significantly reduced engagement with misleading tweets on X/Twitter. Rather, our findings suggest that Community Notes might be too slow to effectively reduce engagement with misinformation in the early (and most viral) stage of diffusion. Our work emphasizes the importance of evaluating fact-checking interventions in the field and offers important implications to enhance crowdsourced fact-checking strategies on social media.

cs.SI↗

arXiv2024

Balancing The Perception of Cheating Detection, Privacy and Fairness: A Mixed-Methods Study of Visual Data Obfuscation in Remote Proctoring

Remote proctoring technology, a cheating-preventive measure, often raises privacy and fairness concerns that may affect test-takers' experiences and the validity of test results. Our study explores how selectively obfuscating information in video recordings can protect test-takers' privacy while ensuring effective and fair cheating detection. Interviews with experts (N=9) identified four key video regions indicative of potential cheating behaviors: the test-taker's face, body, background and the presence of individuals in the background. Experts recommended specific obfuscation methods for each region based on privacy significance and cheating behavior frequency, ranging from conventional blurring to advanced methods like replacement with deepfake, 3D avatars and silhouetting. We then conducted a vignette experiment with potential test-takers (N=259, non-experts) to evaluate their perceptions of cheating detection, visual privacy and fairness, using descriptions and examples of still images for each expert-recommended combination of video regions and obfuscation methods. Our results indicate that the effectiveness of obfuscation methods varies by region. Tailoring remote proctoring with region-specific advanced obfuscation methods can improve the perceptions of privacy and fairness compared to the conventional methods, though it may decrease perceived information sufficiency for detecting cheating. However, non-experts preferred conventional blurring for videos they were more willing to share, highlighting a gap between the perceived effectiveness of the advanced obfuscation methods and their practical acceptance. This study contributes to the field of user-centered privacy by suggesting promising directions to address current remote proctoring challenges and guiding future research.

cs.HC↗