Fernando P. Santos

Dimitris Michailidis, Sennay Ghebreab, Fernando P. Santos

We tackle the Metro Network Expansion Problem (MNEP), a subset of the Transport Network Design Problem (TNDP), which focuses on expanding metro systems to satisfy travel demand. Traditional methods rely on exact and heuristic approaches that require expert-defined constraints to reduce the search space. Recently, deep reinforcement learning (Deep RL) has emerged due to its effectiveness in complex sequential decision-making processes-it remains, however, computationally expensive, environmentally costly, and requires additional engineering to interpret. We show that MNEP problems are small enough to not require Deep RL methods. Reformulating the MNEP as a Non-Markovian Rewards Decision Process (NMRDP), we use tabular RL to achieve similar performance with significantly fewer training episodes, additionally offering greater interpretability. Additionally, we incorporate social equity criteria into the reward functions, focusing on efficiency and fairness, highlighting the versatility of our method. Evaluated in real-world settings-Xi'an and Amsterdam-our method reduces total episodes by a factor of 18 and total carbon emissions by a factor of 12 on average, while remaining competitive with Deep RL. This approach offers a replicable, modular, interpretable, and resource-efficient solution with potential applications to other combinatorial optimization problems.

António Góis, Mehrnaz Mofakhami, Fernando P. Santos et al.

Agents often have individual goals which depend on a group's actions. If agents trust a forecast of collective action and adapt strategically, such prediction can influence outcomes non-trivially, resulting in a form of performative prediction. This effect is ubiquitous in scenarios ranging from pandemic predictions to election polls, but existing work has ignored interdependencies among predicted agents. As a first step in this direction, we study a collective risk dilemma where agents dynamically decide whether to trust predictions based on past accuracy. As predictions shape collective outcomes, social welfare arises naturally as a metric of concern. We explore the resulting interplay between accuracy and welfare, and demonstrate that searching for stable accurate predictions can minimize social welfare with high probability in our setting. By assuming knowledge of a Bayesian agent behavior model, we then show how to achieve better trade-offs and use them for mechanism design.

Dimitris Michailidis, Sennay Ghebreab, Fernando P. Santos

Who benefits from expanding transport networks? While designed to improve mobility, such interventions can also create inequality. In this paper, we show that disparities arise not only from the structure of the network itself but also from differences in how commuters adapt to it. We model commuters as reinforcement learning agents who adapt their travel choices at different learning rates, reflecting unequal access to resources and information. To capture potential efficiency-fairness tradeoffs, we introduce the Price of Learning (PoL), a measure of inefficiency during learning. We analyze both a stylized network -- inspired in the well-known Braess's paradox, yet with two-source nodes -- and an abstraction of a real-world metro system (Amsterdam). Our simulations show that network expansions can simultaneously increase efficiency and amplify inequality, especially when faster learners disproportionately benefit from new routes before others adapt. These results highlight that transport policies must account not only for equilibrium outcomes but also for the heterogeneous ways commuters adapt, since both shape the balance between efficiency and fairness.

Adeela Bashir, Zhao Song, Ndidi Bianca Ogbo et al.

AI safety is an increasingly urgent concern as the capabilities and adoption of AI systems grow. Existing evolutionary models of AI governance have primarily examined incentives for safe development and effective regulation, typically representing users' trust as a one-shot adoption choice rather than as a dynamic, evolving process shaped by repeated interactions. We instead model trust as reduced monitoring in a repeated, asymmetric interaction between users and AI developers, where checking AI behaviour is costly. Using evolutionary game theory, we study how user trust strategies and developer choices between safe (compliant) and unsafe (non-compliant) AI co-evolve under different levels of monitoring cost and institutional regimes. We complement the infinite-population replicator analysis with stochastic finite-population dynamics and reinforcement learning (Q-learning) simulations. Across these approaches, we find three robust long-run regimes: no adoption with unsafe development, unsafe but widely adopted systems, and safe systems that are widely adopted. Only the last is desirable, and it arises when penalties for unsafe behaviour exceed the extra cost of safety and users can still afford to monitor at least occasionally. Our results formally support governance proposals that emphasise transparency, low-cost monitoring, and meaningful sanctions, and they show that neither regulation alone nor blind user trust is sufficient to prevent evolutionary drift towards unsafe or low-adoption outcomes.

Nataliya Balabanova, Adeela Bashir, Paolo Bova et al.

This paper investigates the complex interplay between AI developers, regulators, users, and the media in fostering trustworthy AI systems. Using evolutionary game theory and large language models (LLMs), we model the strategic interactions among these actors under different regulatory regimes. The research explores two key mechanisms for achieving responsible governance, safe AI development and adoption of safe AI: incentivising effective regulation through media reporting, and conditioning user trust on commentariats' recommendation. The findings highlight the crucial role of the media in providing information to users, potentially acting as a form of "soft" regulation by investigating developers or regulators, as a substitute to institutional AI regulation (which is still absent in many regions). Both game-theoretic analysis and LLM-based simulations reveal conditions under which effective regulation and trustworthy AI development emerge, emphasising the importance of considering the influence of different regulatory regimes from an evolutionary game-theoretic perspective. The study concludes that effective governance requires managing incentives and costs for high quality commentaries.

Gabriella D. Franco, Marta C. Couto, Vítor V. Vasconcelos et al.

Recommendation algorithms, used in online social networks, shape interactions between users. In particular, link-recommendation algorithms suggest new connections and affect how individuals interact and exchange information. These algorithms' efficacy relies on key mechanisms governing the creation of social ties, such as triadic closure and homophily. The first is achieved through structural similarity and represents a heightened chance of recommending users to one another given mutual friends; the second is related to opinion similarity and conveys an increased chance of recommending a connection given similar individual characteristics. These two mechanisms jointly shape the evolution of social networks and behaviors unfolding over them. Their combined effect on the co-evolution of opinion and structure dynamics remains, however, poorly understood. Here, we study how social networks and opinions co-evolve given the joint effect of rewiring based on opinion and structural similarity. We show that both similarity metrics lead to polarized states, but differ in how they impact network fragmentation and opinion diversity. While strongly relying on opinion similarity leads to a higher variation of opinion, rewiring via network similarity leads to a larger number of (dis)connected components, resulting in fragmented networks that lean towards one of the signed opinions. Under strong homophilic settings, introducing a weak dependence on structural similarity prevents network fragmentation and favors moderate opinions. This work can inform the design of new recommender algorithms that explicitly account for interacting social and recommendation mechanisms, with the potential to foster moderate opinion coexistence even in inherently polarizing settings.

Alessio Buscemi, Daniele Proverbio, Paolo Bova et al.

There is general agreement that fostering trust and cooperation within the AI development ecosystem is essential to promote the adoption of trustworthy AI systems. By embedding Large Language Model (LLM) agents within an evolutionary game-theoretic framework, this paper investigates the complex interplay between AI developers, regulators and users, modelling their strategic choices under different regulatory scenarios. Evolutionary game theory (EGT) is used to quantitatively model the dilemmas faced by each actor, and LLMs provide additional degrees of complexity and nuances and enable repeated games and incorporation of personality traits. Our research identifies emerging behaviours of strategic AI agents, which tend to adopt more "pessimistic" (not trusting and defective) stances than pure game-theoretic agents. We observe that, in case of full trust by users, incentives are effective to promote effective regulation; however, conditional trust may deteriorate the "social pact". Establishing a virtuous feedback between users' trust and regulators' reputation thus appears to be key to nudge developers towards creating safe AI. However, the level at which this trust emerges may depend on the specific LLM used for testing. Our results thus provide guidance for AI regulation systems, and help predict the outcome of strategic LLM agents, should they be used to aid regulation itself.

Alexandre S. Pires, Laurens Samson, Sennay Ghebreab et al.

Humans increasingly rely on large language models (LLMs) to support decisions in social settings. Previous work suggests that such tools shape people's moral and political judgements. However, the long-term implications of LLM-based social decision-making remain unknown. How will human cooperation be affected when the assessment of social interactions relies on language models? This is a pressing question, as human cooperation is often driven by indirect reciprocity, reputations, and the capacity to judge interactions of others. Here, we assess how state-of-the-art LLMs judge cooperative actions. We provide 21 different LLMs with an extensive set of examples where individuals cooperate -- or refuse cooperating -- in a range of social contexts, and ask how these interactions should be judged. Furthermore, through an evolutionary game-theoretical model, we evaluate cooperation dynamics in populations where the extracted LLM-driven judgements prevail, assessing the long-term impact of LLMs on human prosociality. We observe a remarkable agreement in evaluating cooperation against good opponents. On the other hand, we notice within- and between-model variance when judging cooperation with ill-reputed individuals. We show that the differences revealed between models can significantly impact the prevalence of cooperation. Finally, we test prompts to steer LLM norms, showing that such interventions can shape LLM judgements, particularly through goal-oriented prompts. Our research connects LLM-based advices and long-term social dynamics, and highlights the need to carefully align LLM norms in order to preserve human cooperation.

Dimitris Michailidis, Willem Röpke, Diederik M. Roijers et al.

Multi-Objective Reinforcement Learning (MORL) aims to learn a set of policies that optimize trade-offs between multiple, often conflicting objectives. MORL is computationally more complex than single-objective RL, particularly as the number of objectives increases. Additionally, when objectives involve the preferences of agents or groups, ensuring fairness is socially desirable. This paper introduces a principled algorithm that incorporates fairness into MORL while improving scalability to many-objective problems. We propose using Lorenz dominance to identify policies with equitable reward distributions and introduce λ-Lorenz dominance to enable flexible fairness preferences. We release a new, large-scale real-world transport planning environment and demonstrate that our method encourages the discovery of fair policies, showing improved scalability in two large cities (Xi'an and Amsterdam). Our methods outperform common multi-objective approaches, particularly in high-dimensional objective spaces.

Henrique Correia da Fonseca, António Fernandes, Zhao Song et al.

When developers of artificial intelligence (AI) products need to decide between profit and safety for the users, they likely choose profit. Untrustworthy AI technology must come packaged with tangible negative consequences. Here, we envisage those consequences as the loss of reputation caused by media coverage of their misdeeds, disseminated to the public. We explore whether media coverage has the potential to push AI creators into the production of safe products, enabling widespread adoption of AI technology. We created artificial populations of self-interested creators and users and studied them through the lens of evolutionary game theory. Our results reveal that media is indeed able to foster cooperation between creators and users, but not always. Cooperation does not evolve if the quality of the information provided by the media is not reliable enough, or if the costs of either accessing media or ensuring safety are too high. By shaping public perception and holding developers accountable, media emerges as a powerful soft regulator -- guiding AI safety even in the absence of formal government oversight.

Marta C. Couto, Flavia Barsotti, Fernando P. Santos

Classification algorithms based on Artificial Intelligence (AI) are nowadays applied in high-stakes decisions in finance, healthcare, criminal justice, or education. Individuals can strategically adapt to the information gathered about classifiers, which in turn may require algorithms to be re-trained. Which collective dynamics will result from users' adaptation and algorithms' retraining? We apply evolutionary game theory to address this question. Our framework provides a mathematically rigorous way of treating the problem of feedback loops between collectives of users and institutions, allowing to test interventions to mitigate the adverse effects of strategic adaptation. As a case study, we consider institutions deploying algorithms for credit lending. We consider several scenarios, each representing different interaction paradigms. When algorithms are not robust against strategic manipulation, we are able to capture previous challenges discussed in the strategic classification literature, whereby users either pay excessive costs to meet the institutions' expectations (leading to high social costs) or game the algorithm (e.g., provide fake information). From this baseline setting, we test the role of improving gaming detection and providing algorithmic recourse. We show that increased detection capabilities reduce social costs and could lead to users' improvement; when perfect classifiers are not feasible (likely to occur in practice), algorithmic recourse can steer the dynamics towards high users' improvement rates. The speed at which the institutions re-adapt to the user's population plays a role in the final outcome. Finally, we explore a scenario where strict institutions provide actionable recourse to their unsuccessful users and observe cycling dynamics so far unnoticed in the literature.

Ramona Merhej, Fernando P. Santos, Francisco S. Melo et al.

Collective risk dilemmas (CRDs) are a class of n-player games that represent societal challenges where groups need to coordinate to avoid the risk of a disastrous outcome. Multi-agent systems incurring such dilemmas face difficulties achieving cooperation and often converge to sub-optimal, risk-dominant solutions where everyone defects. In this paper we investigate the consequences of risk diversity in groups of agents learning to play CRDs. We find that risk diversity places new challenges to cooperation that are not observed in homogeneous groups. We show that increasing risk diversity significantly reduces overall cooperation and hinders collective target achievement. It leads to asymmetrical changes in agents' policies -- i.e. the increase in contributions from individuals at high risk is unable to compensate for the decrease in contributions from individuals at low risk -- which overall reduces the total contributions in a population. When comparing RL behaviors to rational individualistic and social behaviors, we find that RL populations converge to fairer contributions among agents. Our results highlight the need for aligning risk perceptions among agents or develop new learning techniques that explicitly account for risk diversity.

Flávio L. Pinheiro, Fernando P. Santos

Designing mechanisms that leverage cooperation between agents has been a long-lasting goal in Multiagent Systems. The task is especially challenging when agents are selfish, lack common goals and face social dilemmas, i.e., situations in which individual interest conflicts with social welfare. Past works explored mechanisms that explain cooperation in biological and social systems, providing important clues for the aim of designing cooperative artificial societies. In particular, several works show that cooperation is able to emerge when specific network structures underlie agents' interactions. Notwithstanding, social dilemmas in which defection is highly tempting still pose challenges concerning the effective sustainability of cooperation. Here we propose a new redistribution mechanism that can be applied in structured populations of agents. Importantly, we show that, when implemented locally (i.e., agents share a fraction of their wealth surplus with their nearest neighbors), redistribution excels in promoting cooperation under regimes where, before, only defection prevailed.