Bias in AI vs. human product recommendations and its impact on consumer trust, attitude toward the product, and purchase intention
Degree
Master of Science in Marketing
Department
Department of Marketing
School
School of Business Studies (SBS)
Date of Submission
Fall 2025-10-15
Supervisor
Dr. Ubedullah Khoso, Assistant Professor, Department of Marketing
Committee Member 1
Dr. Lakhi Mohammad Khosa, Examiner-I, Dean & Professor, SZABIST, Karachi
Committee Member 2
Dr. Ghazanfar Ali Abbasi, Examiner-II, Assistant Professor, King Fahd University, Saudi Arabia
Committee Member 3
Dr. Ashar Saleem, Director Graduate Programs SBS, Institute of Business Administration (IBA), Karachi
Keywords
Product Recommendations, Human vs. AI, Consumer Trust, Purchase Intention, Attitude Towards Products, Persuasion Knowledge
Abstract
Purpose – In this study, we will focus on examining the impact of biased versus unbiased product recommendations originating from two different sources (human expert and AI) on how consumers’ trust is shaped leading towards purchase intentions and attitude toward the product. The study also explores how persuasion knowledge moderates these relationships. For the research, we will focus on Millennial and Generation Z consumers purchasing a functional, high- involvement product: the NextGen Pro smartphone.
Design/methodology/approach – We conducted two separate studies, each with a total of 200 participants, where a between-subjects experimental designs was employed to test out the hypothesis. We compared the impact of biased and unbiased recommendations from human expert and AI algorithms on consumer trust, purchase intention, and attitude towards products. The focus was on a functional product category i.e. a fictitious smartphone brand NextGen Pro.
Findings – Across the two studies, findings demonstrate that consumer trust plays a crucial role in influencing how consumers form attitude towards product and their intention to purchase a product in response to the recommendations. In Study 1, biased recommendations (particularly from AI) led to reduced consumer trust, while unbiased recommendations enhanced trust regardless of source. Study 2 revealed that persuasion knowledge moderated these effects, with highly aware consumers showing greater skepticism toward biased messages, whether they be from AI or human sources.
Originality/value – The study is among of the first to explore how bias in recommendations influences consumer trust, attitude formation towards a certain product and purchase intention across a functional product, specifically among Millennials and Gen Z cohorts. It also identifies persuasion knowledge as a key moderator, demonstrating how consumers’ perception of persuasive intent shapes the impact of biased recommendations on decision-making.
Submission Type
Thesis
Document Type
Restricted Access
The full text of this document is only accessible to authorized users.
