Retail in AI: An ethical dilemma

Publish date:

Consumers expect companies to respect their data when implementing AI.

It is the ultimate shopping trip of the future. A chatbot welcomes you into a store and starts personalizing your experience, while smart digital signage displays product summaries and reviews. The smart fitting room recognizes products, makes recommendations, and calls for assistance when you need it.

This view of a near-future retail experience powered by artificial intelligence (AI) and machine learning (ML) is exciting, but it must be built on an ethical backbone. Without that, the technology will negatively impact the customer experience – and your brand. Customers will be vocal about the misuse of AI, ML, and personalization tactics if they have a bad experience.

According to the new Capgemini Research Institute report Ethics in AI, consumer trust is a major consideration for any AI or ML project. Three in five consumers who perceive an artificial-intelligence interaction as ethical place greater trust in the company, spread positive word of mouth, and are more loyal.

That means ethics should be a major component of any AI or ML strategy, but companies may not be giving it the attention it needs. Only 23% of executives have confidence and insight that their systems run in an ethical manner. Nine out of 10 executives say they have witnessed misuse of AI.

For example, bias is often a result of the data fed into the system. Historical data is not devoid of bias, as it just shows what consumers did in the past. That means depending solely on historical data and derived patterns is not enough. Socio-economic context and expressed customer preferences, among other things, need to be considered.

AI and ML are powerful technologies and companies should deploy them, but building an ethics plan must be part of the process. And it is not too late, even if you have already started down this path. Executives need to lay the foundational practices and processes so customer-facing teams can deploy AI and ML ethically for users. We recommend three stages:

  • Lay strong building blocks for ethical AI: From developing a code of conduct, to creating structures and accountability, to good data-management practices, develop a framework that provides guidance for the entire company.
  • Develop ethics-by-design into systems: Build a diverse team that is aware and empowered. Create user-friendly systems, making sure there are checks and balances to mitigate bias.
  • Deploy these systems ethically: Make sure you educate all of your users and then empower them to work within the ethics framework. If you build an ethics plan, see that it is implemented across the organization.

Ethical AI interactions drive consumer trust and satisfaction. In fact, the CRI research report found that companies that are seen as using AI and ML ethically have a 44-point Net Promoter Score (NPS) advantage over those seen as doing the opposite. For tech-savvy users, it is even higher, up to 84 points. In the retail environment, AI and ML can have a significant impact on revenue.

Previous research has shown that nearly three in four consumers are aware of having interactions enabled by AI and they see the benefits: greater control, 24/7 availability, and convenience. But the move to artificial intelligence and machine learning is a critical juncture, and ethical concerns need to be addressed.

Related Posts

Financial Services

Robo advisors and the Dutch banks: Strategically ignoring the hype or waiting for maturity?

Erdem Tekin
Date icon June 3, 2021

Robo advisors come in different shapes and sizes, making a single definition somewhat...


Pioneering intelligent banking

Elias Ghanem
Date icon March 2, 2021

Eighty-five percent of banking executives stated that AI and AI-powered automation would be...


Acing the returns flow: How to balance costs and value in the last-mile of returned goods?

Wendy Coster – Haas
Date icon March 2, 2021

The main challenges of companies that are dealing with the ever-increasing amount of returns...