Cristian Consonni

Marco Minici, Cristian Consonni, Federico Cinus et al.

Large Language Models (LLMs) increasingly act as gateways to web content, shaping how millions of users encounter online information. Unlike traditional search engines, whose retrieval and ranking mechanisms are well studied, the selection processes of web-connected LLMs add layers of opacity to how answers are generated. By determining which news outlets users see, these systems can influence public opinion, reinforce echo chambers, and pose risks to civic discourse and public trust. This work extends two decades of research in algorithmic auditing to examine how LLMs function as news engines. We present the first audit comparing three leading agents, GPT-4o-Mini, Claude-3.7-Sonnet, and Gemini-2.0-Flash, against Google News, asking: \textit{How do LLMs differ from traditional aggregators in the diversity, ideology, and reliability of the media they expose to users?} Across 24 global topics, we find that, compared to Google News, LLMs surface significantly fewer unique outlets and allocate attention more unevenly. In the same way, GPT-4o-Mini emphasizes more factual and right-leaning sources; Claude-3.7-Sonnet favors institutional and civil-society domains and slightly amplifies right-leaning exposure; and Gemini-2.0-Flash exhibits a modest left-leaning tilt without significant changes in factuality. These patterns remain robust under prompt variations and alternative reliability benchmarks. Together, our findings show that LLMs already enact \textit{agentic editorial policies}, curating information in ways that diverge from conventional aggregators. Understanding and governing their emerging editorial power will be critical for ensuring transparency, pluralism, and trust in digital information ecosystems.

Giovanni De Toni, Cristian Consonni, Erasmo Purificato et al.

Recommendation systems have become central gatekeepers of online information, shaping user behaviour across a wide range of activities. In response, users increasingly organize and coordinate to steer algorithmic outcomes toward diverse goals, such as promoting relevant content or limiting harmful material, relying on platform affordances -- such as likes, reviews, or ratings. While these mechanisms can serve beneficial purposes, they can also be leveraged for adversarial manipulation, particularly in systems where such feedback directly informs safety guarantees. In this paper, we study this vulnerability in recently proposed risk-controlling recommender systems, which use binary user feedback (e.g., "Not Interested") to provably limit exposure to unwanted content via conformal risk control. We empirically demonstrate that their reliance on aggregate feedback signals makes them inherently susceptible to coordinated adversarial user behaviour. Using data from a large-scale online video-sharing platform, we show that a small coordinated group (comprising only 1% of the user population) can induce up to a 20% degradation in nDCG for non-adversarial users by exploiting the affordances provided by risk-controlling recommender systems. We evaluate simple, realistic attack strategies that require little to no knowledge of the underlying recommendation algorithm and find that, while coordinated users can significantly harm overall recommendation quality, they cannot selectively suppress specific content groups through reporting alone. Finally, we propose a mitigation strategy that shifts guarantees from the group level to the user level, showing empirically how it can reduce the impact of adversarial coordinated behaviour while ensuring personalized safety for individuals.

Giovanni De Toni, Erasmo Purificato, Emilia Gómez et al.

Recommenders are significantly shaping online information consumption. While effective at personalizing content, these systems increasingly face criticism for propagating irrelevant, unwanted, and even harmful recommendations. Such content degrades user satisfaction and contributes to significant societal issues, including misinformation, radicalization, and erosion of user trust. Although platforms offer mechanisms to mitigate exposure to undesired content, these mechanisms are often insufficiently effective and slow to adapt to users' feedback. This paper introduces an intuitive, model-agnostic, and distribution-free method that uses conformal risk control to provably bound unwanted content in personalized recommendations by leveraging simple binary feedback on items. We also address a limitation of traditional conformal risk control approaches, i.e., the fact that the recommender can provide a smaller set of recommended items, by leveraging implicit feedback on consumed items to expand the recommendation set while ensuring robust risk mitigation. Our experimental evaluation on data coming from a popular online video-sharing platform demonstrates that our approach ensures an effective and controllable reduction of unwanted recommendations with minimal effort. The source code is available here: https://github.com/geektoni/mitigating-harm-recsys.

Cristian Consonni, Silvia Basile, Matteo Manca et al.

Mobility and transport, by their nature, involve crowds and require the coordination of multiple stakeholders - such as policy-makers, planners, transport operators, and the travelers themselves. However, traditional approaches have been focused on time savings, proposing to users solutions that include the shortest or fastest paths. We argue that this approach towards travel time value is not centered on a traveler's perspective. To date, very few works have mined data from crowds of travelers to test the efficacy and efficiency of novel mobility paradigms. In this paper, we build upon a different paradigm of worthwhile time in which travelers can use their travel time for other activities; we present a new dataset, which contains data about travelers and their journeys, collected from a dedicated mobile application. Each trip contains multi-faceted information: from the transport mode, through its evaluation, to the positive and negative experience factors. To showcase this new dataset's potential, we also present a use case, which compares corresponding trip legs with different transport modes, studying experience factors that negatively impact users using cycling and public transport as alternatives to cars. We conclude by discussing other application domains and research opportunities enabled by the dataset.

Giovanni De Toni, Cristian Consonni, Alberto Montresor

Influenza is an acute respiratory seasonal disease that affects millions of people worldwide and causes thousands of deaths in Europe alone. Being able to estimate in a fast and reliable way the impact of an illness on a given country is essential to plan and organize effective countermeasures, which is now possible by leveraging unconventional data sources like web searches and visits. In this study, we show the feasibility of exploiting information about Wikipedia's page views of a selected group of articles and machine learning models to obtain accurate estimates of influenza-like illnesses incidence in four European countries: Italy, Germany, Belgium, and the Netherlands. We propose a novel language-agnostic method, based on two algorithms, Personalized PageRank and CycleRank, to automatically select the most relevant Wikipedia pages to be monitored without the need for expert supervision. We then show how our model is able to reach state-of-the-art results by comparing it with previous solutions.