[Paper Review] Incentivizing the Dissemination of Truth Versus Fake News in Social Networks

Abbas Ehsanfar, Mo Mansouri|arXiv (Cornell University)|Apr 7, 2018

Opinion Dynamics and Social Influence2 references3 citations

TL;DR

This paper models the dissemination of truth versus fake news in social networks as a volunteer's dilemma, where users collectively contribute to verifying news. It demonstrates that shared rewards through news subscriptions can incentivize truthful information dissemination, reducing fake news dominance by aligning individual incentives with collective truth-seeking in networked systems.

ABSTRACT

The concept of truth, as a public good is the production of a collective understanding, which emerges from a complex network of social interactions. The recent impact of social networks on shaping the perception of truth in political arena shows how such perception is corroborated and established by the online users, collectively. However, investigative journalism for discovering truth is a costly option, given the vast spectrum of online information. In some cases, both journalist and online users choose not to investigate the authenticity of the news they receive, because they assume other actors of the network had carried the cost of validation. Therefore, the new phenomenon of "fake news" has emerged within the context of social networks. The online social networks, similarly to System of Systems, cause emergent properties, which makes authentication processes difficult, given availability of multiple sources. In this study, we show how this conflict can be modeled as a volunteer's dilemma. We also show how the public contribution through news subscription (shared rewards) can impact the dominance of truth over fake news in the network.

Motivation & Objective

• To address the challenge of fake news proliferation in social networks, where users often fail to verify information due to free-riding on others' efforts.
• To analyze how collective verification of news can emerge as a public good in networked environments with decentralized trust.
• To investigate the role of shared rewards (e.g., subscriptions) in incentivizing users to verify and share truthful content.
• To model the dynamics of truth and fake news dissemination as a system-level emergent behavior influenced by individual incentives.
• To evaluate how system design—particularly reward-sharing mechanisms—can steer network behavior toward truth-telling.

Proposed method

• Models the problem as a volunteer's dilemma, where each user decides whether to invest effort in verifying news or rely on others.
• Introduces a shared reward mechanism where subscribers collectively fund verification efforts, distributing benefits across contributors.
• Uses game-theoretic modeling to analyze equilibrium behavior in networks with heterogeneous user incentives.
• Analyzes network-level outcomes by simulating the impact of varying reward-sharing structures on truth propagation.
• Applies System of Systems (SoS) engineering principles to understand emergent behaviors in interconnected social networks.
• Evaluates the stability and dominance of truth versus fake news under different incentive structures using analytical and simulation-based methods.

Experimental results

Research questions

• RQ1 How does the volunteer's dilemma framework explain the failure to verify news in social networks?
• RQ2 What role do shared rewards play in incentivizing users to verify and disseminate truthful information?
• RQ3 Under what conditions does truthful content dominate over fake news in a networked environment?
• RQ4 How do network structure and reward-sharing mechanisms affect the emergence of collective truth-seeking behavior?
• RQ5 Can system-level incentives be designed to reduce free-riding and increase the dissemination of verified information?

Key findings

• The volunteer's dilemma framework accurately captures the disincentive for individual users to verify news due to free-riding on others' efforts.
• Shared rewards through news subscriptions significantly increase the likelihood of truthful content being disseminated over fake news.
• Network-wide truth dominance emerges when verification costs are shared and users perceive a fair return on their verification efforts.
• The model shows that even moderate levels of shared rewards can shift equilibrium behavior toward higher verification rates.
• System-level incentives, particularly when aligned with collective benefit, can counteract the natural tendency toward fake news spread.
• Emergent behaviors in social networks—such as truth propagation—can be steered through strategic reward design, even in decentralized systems.