,
Associate Professor & Programme Coordinator MBA (AIDS), Faculty of Management, SRM Institute of Science and Technology, Kattankulathur, Tamil Nadu, India
,
Professor & Dean, Faculty of Management, SRM Institute of Science and Technology, Kattankulathur, Tamil Nadu, India
,
Professor, Department of Management, Sharda School of Business Studies, Sharda University, Greater Noida, Uttar Pradesh, India
Assistant Professor, Faculty of Management, SRM institute of Science and Technology, Kattankulathur, Tamil Nadu, India
The growing popularity of AI-generated virtual influencers on platforms like Instagram, TikTok, and YouTube has brought about changes in digital marketing through scalable and controllable promotion strategies for brands. Unfortunately, there is still a lack of quantitative models to help compare and measure consumer perceptions about AI-generated influencers and human influencers, especially in terms of trustworthiness and engagement dynamics. This research presents a machine learning-based framework to be used in the systematic analysis and comparison of consumer perceptions towards AI and human influencers in terms of brand promotion. The main goal of this research is to come up with a single model that measures perception differences between the two groups through sentiment analysis and engagement. This methodology uses a combination of data consisting of 15,000 social media posts and survey responses based on the Likert scale (1–5) along with engagement metrics such as likes, shares, and comments. Data preprocessing entails NLP-based cleaning, tokenization, normalization, and feature scaling. The proposed architecture involves the use of BERT based on the Transformer model for sentiment analysis, CNN for feature extraction for engagements, and Random Forest for classification purposes, where the fusion layer is used to develop unified perception metrics. The experimental findings show that the classification accuracy of the proposed framework is 96.9%, which outperforms baseline models such as SVM (88.2%), LSTM (92.5%), and CNN (93.8%). It is also shown that the engagement prediction improvement is between 12% and 18% over the baselines, with a Perception Divergence Index (PDI) value of 0.42, which shows there is a divergence between AI-generated and human influencers' perceptions of the consumers. The research findings show that AI-generated influencers produce high levels of engagement driven by novelty, while human influencers have high consumer trust.
This is an open access article distributed under the Creative Commons Attribution Non-Commercial License (CC BY-NC) License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The statements, opinions and data contained in the journal are solely those of the individual authors and contributors and not of the publisher and the editor(s). We stay neutral with regard to jurisdictional claims in published maps and institutional affiliations.
