The potential of AI in empowering consumers

Live Mint, January 09, 2019

By Pradeep S Mehta

In April 2018, the department of economic development, Dubai, launched a “Smart Protection” service, which adopts Artificial Intelligence (AI) to respond efficiently to consumer queries and resolve their complaints. Through an app called Dubai Consumer, the service engages consumers in a direct dialogue to gather information and, within a few minutes, issues an “empowerment letter” stating details of complaint and instructions to the retailer to resolve relevant grievance within a pre-specified time frame, failing which the retailer risks attracting fines.

It was reported in November that retailers comply with instructions in empowerment letters in more than 90% cases. The service has been trained to handle grievances in more than 12 sectors and understands more than 40 laws and regulations relating to consumer protection.

Imagine the utility of such a service in our country, wherein the standard of customer support and grievance redressal is abysmally poor. Consumer courts suffer from infrastructure and capacity constraints and even simple matters drag on for years. Our experience of consumer grievance handling reveals stark information asymmetry among consumers and high opportunity cost of filing and pursuing complaints. The use of AI in addressing these constraints can go a long way in empowering consumers.

Alas, the discourse on AI has been captured by concepts such as bots surpassing human intelligence, ethical use of AI, and algorithmic bias. However, the potential of AI in empowering consumers must not be lost sight of. A recent paper by Giuseppe Contissa et al has done a huge service by bringing this issue into the limelight.

The need to empower consumers is now more than ever, given the unprecedented growth of the digital economy and global interconnectedness. Fortunately, awareness about consumer rights, especially privacy and data protection, has increased manifold owing to the Puttaswamy judgment and discussions around the draft personal data protection bill. Despite being aware, consumers lack adequate tools to protect their privacy and data. Measures hitherto suggested, such as increasing transparency by providing more information, may result in information overload and compound the asymmetry.

AI can help here as well. Tools like CLAUDETTE can be used for automatic detection of potentially unfair clauses in contracts and terms of service, and are being improved to assess compliance with data protection regulation. PriBot and Polisis provide AI-powered summaries and a free-form question answering system for privacy policies. However, these tools are being developed further, primarily interact in English, and require users to approach them to obtain information. These constraints may limit scalability in India.

Thus, there is a need of a bottom-up approach to leverage AI for Indian consumers. Can AI-powered privacy labels be designed that can present the most important terms and conditions, depending on consumers’ likely profile and behaviour, in a format preferred by them, thus facilitating informed decision making? Can privacy policy reviewing tools automatically alert regulatory agencies of unfair terms to overcome consumers’ lackadaisical attitude in enforcing their rights and filing grievances? Customized solutions contextualized to meet specific needs can go a long way in empowering Indian consumers.

The extraordinary progress in the digital economy coupled with smartphone penetration in the country has put consumers in the midst of an information deluge, making it difficult for them to compare products intelligently. AI-empowered personal assistants can help them navigate through exponentially growing data, reduce search and transaction costs, overcome biases, and enable more sophisticated choices. These tools may also help consumers assess the trustworthiness and helpfulness of online consumer reviews by analysing sentiment and polarity.

The day is not far when AI does not merely assist consumers in decision making but takes decisions on behalf of consumers. Michal S. Gal et al suggest that the next generation of e-commerce will be conducted by digital agents, based on algorithms that will not only make purchase recommendations but will also predict what consumers want, make purchase decisions, negotiate and execute transactions for consumers, and even automatically form coalitions of buyers to enjoy better terms, thereby replacing human decision-making.

Unsurprisingly, despite its potential in empowering consumers, the increasing role of AI in the lives of consumers is fraught with conceptual, moral, ethical and regulatory dilemmas. India’s sociocultural and economic diversity may pose challenges and delay the mass utilisation of AI, but such time needs to be judiciously utilised in designing principles and frameworks to address related dilemmas.

The time to discuss such principles and frameworks is not in the future, but now. India is modernizing its regulatory frameworks on privacy and data protection, consumer protection, e-commerce. We cannot let go of this opportunity and need to be bold and forward looking to move collectively towards consumer-empowering AI.

Pradeep S. Mehta is secretary general of CUTS International

Udai Mehta and Amol Kulkarni of CUTS contributed to this article.

This news item can also be viewed at: