Authentic reviews, whether positive or negative, serve as a valuable resource by sharing genuine experiences. In contrast, fake reviews distort the marketplace by depriving consumers of accurate information and swaying their decisions with misleading endorsements.
Authors: Mayank Mishra and Salil Ahuja
Published: July 27, 2023 in Hindustan Times
While there have been several discussions on how this may impact our society, we specifically examine the challenges that may emerge in digital commerce. Within a few months of generative AI’s rollout for common use, we’re witnessing the early signs of AI tools, like ChatGPT, being misused to power bots, generate fake reviews, and saturate the web with subpar content. The products and services listed on online marketplaces are already witnessing reviews that begin with the telltale words, “As an AI language model…”, which clearly implies that these reviews have been written with the help of generative AI.
This is a significant threat to trust in digital platforms because when it comes to making purchasing decisions, the opinions of consumers, who have used a particular product, hold significant sway. According to a survey by an SEO platform, a staggering 76% of consumers regularly consider reviews before making a product choice. Furthermore, nearly 75% admit that a positive review is the key factor influencing their perception of the product. However, the proliferation of fake reviews jeopardises consumer trust in online reviews as a whole. Authentic reviews, whether positive or negative, serve as a valuable resource by sharing genuine experiences. In contrast, fake reviews distort the marketplace by depriving consumers of accurate information and swaying their decisions with misleading endorsements. Fake reviews also hold the power to either damage or bolster a business’s reputation, with far-reaching consequences. Research conducted by Harvard Professor Michael Luca suggests that a mere one-star change in a company’s Yelp rating can substantially impact its revenue by 5-9%.
In the early days of digital commerce, there wasn’t significant recognition of the need to regulate platforms to ensure genuineness of reviews. Unfortunately, this led to the rise of paid-review farms, which exploit businesses seeking to artificially boost their ratings on review platforms. A different study by Luca highlighted that in 2015, fake reviews could be brought for as little as 25 cents per review, indicating that the cost of posting fake reviews was much lower than the benefit one could reap out of them.
Recognising the severity of the problem, the European Commission (EU) and national consumer protection authorities conducted an extensive EU-wide website screening to assess the credibility of consumer reviews. The results revealed deep doubts about the authenticity of reviews on the majority of the analysed websites. In India, the proliferation of fake and misleading reviews also violates consumers’ right to be informed, as enshrined in the Consumer Protection Act, 2019. The Department of Consumer Affairs therefore took notice after the release of the EU report and promptly invited key stakeholders to address the issue.
As a result of these discussions, the Department in collaboration with the Bureau of Indian Standards (BIS), introduced voluntary standards for online reviews in November 2022. However, when these standards were created, they did not proactively account for the generative abilities of AI. It is now crucial to consider the ongoing AI boom and its implications on the industry.
AI is a dynamic field, and regulating fake reviews online requires the support and collaboration of industry stakeholders. The government must encourage innovation by the private sector in collaboration with experts and consumer groups to address this problem at scale as proliferation of fake reviews is likely to increase given the short time required to generate new content. Online marketplaces also have market incentives to address this problem, as they are well aware that fake reviews undermine trust and credibility, leading to decreased user engagement and ultimately people’s reluctance to use their platforms. We must therefore also build systems so that platforms collaborate with each other to identify common bad actors who are operating across online marketplaces.
Further, in India, initiatives such as the Open Network for Digital Commerce (ONDC) which are encouraging interoperability could also play a potential role in building systems of cooperation to check fake reviews. Fixing accountability of fake reviews and law enforcement mechanisms may also need to be revisited as AI’s impact on this aspect becomes clearer.
As the AI revolution unfolds, our ability to preserve the integrity of digital commerce becomes paramount to ensure continued trust in digital markets. We should take proactive steps to regulate and combat the proliferation of fake reviews to enable informed decision-making and protect online commerce from AI-powered manipulative practices.
Mayank Mishra is Manager, Public Policy and Salil Ahuja is Associate, Public Policy at TQH Consulting.