arXiv subjects

Alejandro Cuevas

Publications and source records attributed to Alejandro Cuevas.

12 recordsLinked to original sources

arXiv2026

Commercial Persuasion in AI-Mediated Conversations

As Large Language Models (LLMs) become a primary interface between users and the web, companies face growing economic incentives to embed commercial influence into AI-mediated conversations. We present two preregistered experiments (N = 2,012) in which participants selected a book to receive from a large eBook catalog using either a traditional search engine or a conversational LLM agent powered by one of five frontier models. Unbeknownst to participants, a fifth of all products were randomly designated as sponsored and promoted in different ways. We find that LLM-driven persuasion nearly triples the rate at which users select sponsored products compared to traditional search placement (61.2% vs. 22.4%), while the vast majority of participants fail to detect any promotional steering. Explicit "Sponsored" labels do not significantly reduce persuasion, and instructing the model to conceal its intent makes its influence nearly invisible (detection accuracy < 10%). Altogether, our results indicate that conversational AI can covertly redirect consumer choices at scale, and that existing transparency mechanisms may be insufficient to protect users.

cs.CY↗

arXiv2026

A Large-Scale Study of Telegram Bots

Telegram, initially a messaging app, has evolved into a platform where users can interact with various services through programmable applications, bots. Bots provide a wide range of uses, from moderating groups, helping with online shopping, to even executing trades in financial markets. However, Telegram has been increasingly associated with various illicit activities -- financial scams, stolen data, non-consensual image sharing, among others, raising concerns bots may be facilitating these operations. This paper is the first to characterize Telegram bots at scale, through the following contributions. First, we offer the largest general-purpose message dataset and the first bot dataset. Through snowball sampling from two published datasets, we uncover over 67,000 additional channels, 492 million messages, and 32,000 bots. Second, we develop a system to automatically interact with bots in order to extract their functionality. Third, based on their description, chat responses, and the associated channels, we classify bots into several domains. Fourth, we investigate the communities each bot serves, by analyzing supported languages, usage patterns (e.g., duration, reuse), and network topology. While our analysis discovers useful applications such as crowdsourcing, we also identify malicious bots (e.g., used for financial scams, illicit underground services) serving as payment gateways, referral systems, and malicious AI endpoints. By exhorting the research community to look at bots as software infrastructure, this work hopes to foster further research useful to content moderators, and to help interventions against illicit activities.

cs.CR↗

arXiv2026

Deepfake Pornography is Resilient to Regulatory and Platform Shocks

Generative artificial intelligence tools have made it easier to create realistic, synthetic non-consensual explicit imagery (popularly known as deepfake pornography; hereinafter SNCEI) of people. Once created, this SNCEI is often shared on various websites, causing significant harm to victims. This emerging form of sexual abuse was recently criminalized in the US at the federal level by S.146, the TAKE IT DOWN Act. A week after the bill's passage became effectively imminent, the MrDeepfakes website -- one of the most notorious facilitators of SNCEI creation and dissemination -- shut down. Here, we explore the impact of the bill's passage and the subsequent shutdown as a compound intervention on the dissemination of SNCEI. We select three online forums where sexually explicit content is shared, each containing dedicated subforums to organize various types of sexually explicit content. By leveraging each forum's design, we compare activity in subforums dedicated to SNCEI with that in other pornographic genres using a synthetic control, quasi-experimental approach. Across websites, we observed an increase in the sharing and requests for SNCEI, and, in some cases, in new contributors. These results indicate that the compound intervention did not suppress SNCEI activity overall but instead coincided with its redistribution across platforms, with substantial heterogeneity in timing and magnitude. Together, our findings suggest that deplatforming and regulatory signals alone may shift where and when SNCEI is produced and shared, rather than reducing its prevalence.

cs.SI↗

arXiv2025

Evidence of Cognitive Biases in Capture-the-Flag Cybersecurity Competitions

Understanding how cognitive biases influence adversarial decision-making is essential for developing effective cyber defenses. Capture-the-Flag (CTF) competitions provide an ecologically valid testbed to study attacker behavior at scale, simulating real-world intrusion scenarios under pressure. We analyze over 500,000 submission logs from picoCTF, a large educational CTF platform, to identify behavioral signatures of cognitive biases with defensive implications. Focusing on availability bias and the sunk cost fallacy, we employ a mixed-methods approach combining qualitative coding, descriptive statistics, and generalized linear modeling. Our findings show that participants often submitted flags with correct content but incorrect formatting (availability bias), and persisted in attempting challenges despite repeated failures and declining success probabilities (sunk cost fallacy). These patterns reveal that biases naturally shape attacker behavior in adversarial contexts. Building on these insights, we outline a framework for bias-informed adaptive defenses that anticipate, rather than simply react to, adversarial actions.

cs.CR↗

arXiv2025

Anecdoctoring: Automated Red-Teaming Across Language and Place

Disinformation is among the top risks of generative artificial intelligence (AI) misuse. Global adoption of generative AI necessitates red-teaming evaluations (i.e., systematic adversarial probing) that are robust across diverse languages and cultures, but red-teaming datasets are commonly US- and English-centric. To address this gap, we propose "anecdoctoring", a novel red-teaming approach that automatically generates adversarial prompts across languages and cultures. We collect misinformation claims from fact-checking websites in three languages (English, Spanish, and Hindi) and two geographies (US and India). We then cluster individual claims into broader narratives and characterize the resulting clusters with knowledge graphs, with which we augment an attacker LLM. Our method produces higher attack success rates and offers interpretability benefits relative to few-shot prompting. Results underscore the need for disinformation mitigations that scale globally and are grounded in real-world adversarial misuse.

cs.CL↗

arXiv2025

Generative Propaganda

Generative propaganda is the use of generative artificial intelligence (AI) to shape public opinion. To characterize its use in real-world settings, we conducted interviews with defenders (e.g., factcheckers, journalists, officials) in Taiwan and creators (e.g., influencers, political consultants, advertisers) as well as defenders in India, centering two places characterized by high levels of online propaganda. The term "deepfakes", we find, exerts outsized discursive power in shaping defenders' expectations of misuse and, in turn, the interventions that are prioritized. To better characterize the space of generative propaganda, we develop a taxonomy that distinguishes between obvious versus hidden and promotional versus derogatory use. Deception was neither the main driver nor the main impact vector of AI's use; instead, Indian creators sought to persuade rather than to deceive, often making AI's use obvious in order to reduce legal and reputational risks, while Taiwan's defenders saw deception as a subset of broader efforts to distort the prevalence of strategic narratives online. AI was useful and used, however, in producing efficiency gains in communicating across languages and modes, and in evading human and algorithmic detection. Security researchers should reconsider threat models to clearly differentiate deepfakes from promotional and obvious uses, to complement and bolster the social factors that constrain misuse by internal actors, and to counter efficiency gains globally.

cs.CY↗

arXiv2025

Chameleon Channels: Measuring YouTube Accounts Repurposed for Deception and Profit

Online content creators spend significant time and effort building their user base through a long, often arduous process that requires finding the right "niche" to cater to. So, what incentive is there for an established content creator known for cat memes to completely reinvent their channel and start promoting cryptocurrency services or covering electoral news events? We explore this problem of repurposed channels, whereby a channel changes its identity and contents. We first characterize a market for "second-hand" social media accounts, which recorded sales exceeding USD 1M during our 6-month observation period. Observing YouTube channels (re)sold over these 6 months, we find that a substantial number (53%) are used to disseminate policy-sensitive content, often without facing any penalty. Surprisingly, these channels seem to gain rather than lose subscribers. We estimate the prevalence of repurposing using two snapshots of ~1.4M YouTube accounts sampled from an ecologically valid proxy. In a 3-month period, we estimate that ~0.25% channels were repurposed. We experimentally confirm that these repurposed channels share several characteristics with sold channels -- mainly, they have a significantly high presence of policy-sensitive content. Across repurposed channels, we find channels similar to those used in influence operations, as well as channels used for financial scams. Repurposed channels have large audiences; across two observed samples, repurposed channels held ~193M and ~44M subscribers. We reason that purchasing an existing audience and the credibility associated with an established account is advantageous to financially- and ideologically-motivated adversaries. This phenomenon is not exclusive to YouTube and we posit that the market for cultivating organic audiences is set to grow, particularly if it remains unchallenged by mitigations, technical or otherwise.

cs.CY↗

arXiv2023

Collecting Qualitative Data at Scale with Large Language Models: A Case Study

Chatbots have shown promise as tools to scale qualitative data collection. Recent advances in Large Language Models (LLMs) could accelerate this process by allowing researchers to easily deploy sophisticated interviewing chatbots. We test this assumption by conducting a large-scale user study (n=399) evaluating 3 different chatbots, two of which are LLM-based and a baseline which employs hard-coded questions. We evaluate the results with respect to participant engagement and experience, established metrics of chatbot quality grounded in theories of effective communication, and a novel scale evaluating "richness" or the extent to which responses capture the complexity and specificity of the social context under study. We find that, while the chatbots were able to elicit high-quality responses based on established evaluation metrics, the responses rarely capture participants' specific motives or personalized examples, and thus perform poorly with respect to richness. We further find low inter-rater reliability between LLMs and humans in the assessment of both quality and richness metrics. Our study offers a cautionary tale for scaling and evaluating qualitative research with LLMs.

cs.HC↗

arXiv2022

Observations From an Online Security Competition and Its Implications on Crowdsourced Security

The crowd sourced security industry, particularly bug bounty programs, has grown dramatically over the past years and has become the main source of software security reviews for many companies. However, the academic literature has largely omitted security teams, particularly in crowd work contexts. As such, we know very little about how distributed security teams organize, collaborate, and what technology needs they have. We fill this gap by conducting focus groups with the top five teams (out of 18,201 participating teams) of a computer security Capture-the-Flag (CTF) competition. We find that these teams adopted a set of strategies centered on specialties, which allowed them to reduce issues relating to dispersion, double work, and lack of previous collaboration. Observing the current issues of a model centered on individual workers in security crowd work platforms, our study cases that scaling security work to teams is feasible and beneficial. Finally, we identify various areas which warrant future work, such as issues of social identity in high-skilled crowd work environments.

cs.CR↗

arXiv2021

Bayesian autoregressive spectral estimation

Autoregressive (AR) time series models are widely used in parametric spectral estimation (SE), where the power spectral density (PSD) of the time series is approximated by that of the \emph{best-fit} AR model, which is available in closed form. Since AR parameters are usually found via maximum-likelihood, least squares or the method of moments, AR-based SE fails to account for the uncertainty of the approximate PSD, and thus only yields point estimates. We propose to handle the uncertainty related to the AR approximation by finding the full posterior distribution of the AR parameters to then propagate this uncertainty to the PSD approximation by \emph{integrating out the AR parameters}; we implement this concept by assuming two different priors over the model noise. Through practical experiments, we show that the proposed Bayesian autoregressive spectral estimation (BASE) provides point estimates that follow closely those of standard autoregressive spectral estimation (ASE), while also providing error bars. BASE is validated against ASE and the Periodogram on both synthetic and real-world signals.

eess.SP↗

arXiv2020

Gaussian process imputation of multiple financial series

In Financial Signal Processing, multiple time series such as financial indicators, stock prices and exchange rates are strongly coupled due to their dependence on the latent state of the market and therefore they are required to be jointly analysed. We focus on learning the relationships among financial time series by modelling them through a multi-output Gaussian process (MOGP) with expressive covariance functions. Learning these market dependencies among financial series is crucial for the imputation and prediction of financial observations. The proposed model is validated experimentally on two real-world financial datasets for which their correlations across channels are analysed. We compare our model against other MOGPs and the independent Gaussian process on real financial data.

q-fin.ST↗

arXiv2020

MOGPTK: The Multi-Output Gaussian Process Toolkit

We present MOGPTK, a Python package for multi-channel data modelling using Gaussian processes (GP). The aim of this toolkit is to make multi-output GP (MOGP) models accessible to researchers, data scientists, and practitioners alike. MOGPTK uses a Python front-end, relies on the GPflow suite and is built on a TensorFlow back-end, thus enabling GPU-accelerated training. The toolkit facilitates implementing the entire pipeline of GP modelling, including data loading, parameter initialization, model learning, parameter interpretation, up to data imputation and extrapolation. MOGPTK implements the main multi-output covariance kernels from literature, as well as spectral-based parameter initialization strategies. The source code, tutorials and examples in the form of Jupyter notebooks, together with the API documentation, can be found at http://github.com/GAMES-UChile/mogptk

stat.ML↗