New AAAI research examines AI’s impact on people, platforms and markets

Researchers affiliated with the Managing Disruption: Analytics, Advanced Digital Technologies and AI (AAAI) initiative at McGill University’s Desautels Faculty of Management are advancing our understanding of how artificial intelligence is reshaping human interactions, online communities and market outcomes. In three recent studies, Professor Elizabeth Han, Professor Sameer Borwankar, Professor Warut Khern-am-nuai, Professor Maxime Cohen and PhD student Jitsama Tanlamai examine important dimensions of AI adoption. Their research explores how people respond to emotionally intelligent AI systems, how online communities adapt when access to generative AI tools is restricted, and whether generative AI can help reduce disparities in housing prices.

Professor Elizabeth Han’s research, “Bots with Empathy: Reactance Against Emotion-Aware AI Agents in Customer Service” investigates whether customer service chatbots should express empathy after service failures. While empathy is generally considered beneficial in human interactions, the study finds that when chatbots express empathy, users may experience psychological reactance, which may stem from the perception that the AI is being intrusive and crossing the boundary between humans and machines. This can reduce the perceptions of the chatbot’s competence and warmth, ultimately lowering customer satisfaction.

The findings highlight the importance of carefully designing emotionally intelligent AI systems rather than assuming that communication strategies effective in human interactions will produce the same results when used by AI.

In “Unraveling the Impact: An Empirical Investigation of ChatGPT’s Exclusion from Stack Overflow,” professors Sameer Borwankar and Warut Khern-am-nuai examine how restricting AI-generated content affects online knowledge-sharing communities. Using data from Stack Overflow and Reddit, along with evidence from Italy’s temporary ChatGPT ban and online experiments, they found that contributors produced longer, more positive and more linguistically complex answers that received greater community approval after ChatGPT-generated content was restricted.

However, participation declined, with fewer questions, answers and first-time contributors. The study highlights a key challenge for organizations regulating generative AI: restrictions may improve the perceived quality of human contributions while reducing user engagement.

In their study, "Generative AI and Price Discrimination in the Housing Market," PhD student Jitsama Tanlamai and professors Warut Khern-am-nuai and Maxime Cohen explore whether generative AI can help reduce longstanding disparities in housing prices. Using data from more than 284,000 U.S. properties, the researchers compare human-generated listing prices with prices generated by generative AI and find that, while AI-generated prices are not entirely free from discrimination, they reduce the pricing gap between comparable homes in white-dominant and minority-dominant neighbourhoods by 8.7%.

These findings contrast with previous research showing that traditional AI models can reproduce or amplify historical biases. The results suggest that generative AI may instead help alleviate certain forms of price discrimination. The study highlights the potential for generative AI to support more equitable decision-making while emphasizing that it is not a complete solution to fairness concerns and that outcomes can remain sensitive to how these systems are used.

---

Analytics, Advanced Digital Technologies and AI Initiative (AAAI)

AAAI aims to promote and support academic research, teaching activities, intra- and inter-university collaborations, and industry and community engagements related to artificial intelligence (AI), analytics, and advanced digital technologies such as blockchain, Internet of things (IoT), and Metaverse. The initiative encourages faculty research that leverages advanced technologies and foster collaborations both inside and outside the faculty, facilitate the inclusion of advanced technologies in classroom and pedagogical practices, and undertake outreach with external communities to both disseminate academic knowledge and obtain practical oriented insights to further advance the objective of this initiative in the faculty.