Ethical aspects of multi-stakeholder recommendation systems

Abstract

This article analyses the ethical aspects of multistakeholder recommendation systems (RSs). Following the most common approach in the literature, we assume a consequentialist framework to introduce the main concepts of multistakeholder recommendation. We then consider three research questions: who are the stakeholders in a RS? How are their interests taken into account when formulating a recommendation? And, what is the scientific paradigm underlying RSs? Our main finding is that multistakeholder RSs (MRSs) are designed and theorised, methodologically, according to neoclassical welfare economics. We consider and reply to some methodological objections to MRSs on this basis, concluding that the multistakeholder approach offers the resources to understand the normative social dimension of RSs.

Leveraging Artificial Intelligence in Marketing for Social Good—An Ethical Perspective

“…On the individual level, exploitation of customer information that AI systems already possess constitutes the optimal (standard) strategy to maximize individual utility by satisfying preferences. Conversely, AI systems’ exploratory recommendations of new alternatives (e.g., sustainable items) might be the strategy with greatest expected utilities on a societal level (Milano et al, 2021 ). The environmental impact and material footprint of consumption (Wiedmann et al, 2015 ) that could be additionally fueled by AI applications in marketing contravene the beneficence principle of promoting well-being of humans and the planet.…”

“…In contrast, these ethical judgments do not coincide on the company and customer levels. That means, AI applications could be beneficent and maleficent at the same time (Milano et al, 2021 ).…”

“…On the company level, recommender systems can discriminate firms by decreasing the variety of products consumers explore and purchase (i.e., sales diversity) and by increasing market share concentration for popular products (Lee & Hosaganar, 2019 ). Besides, the rise of multisided e-commerce platforms deploying AI-generated recommendations can endanger traditional retailers that do not participate in these platforms by undermining their business models and impeding market access (Milano et al, 2021 ).…”

“…Based on comprehensive purchase (history) and user data, recommendations could nudge consumers to sustainable products or at least offer sustainable alternatives to conventional products. That is, recommender systems should balance exploration of (new sustainable) items and exploitation of existing preferences for their recommendations (Milano et al, 2021 ). In this way, diversity of product offerings could be increased in terms of environmental friendliness, material footprint, and sustainability (e.g., fast- vs. slow fashion items).…”

“…On the customer level, discrimination can arise from customer prioritization based on demographic and economic factors (e.g., Libai et al, 2020 ) and targeting (e.g., Matz & Netzer, 2017 ) or alienation (e.g., Puntoni et al, 2021 ) of vulnerable consumer groups. On the company level, market share concentration through AI-enabled e-commerce platforms (e.g., Lee & Hosaganar, 2019 ) and unequal representation on them (e.g., Milano et al, 2021 ) can disadvantage some companies while privileging others. Such discriminatory treatments can reinforce and exacerbate existing economic and societal inequalities.…”