Artificial intelligence is intended to make online trading a more personal and easier experience for customers. In addition to opportunities, there are also risks for customers.
Which running shoes are best for uneven terrain? What is the difference between a French press pot and a filter coffee machine? Which toys are particularly suitable for six-year-olds? If you have certain questions when shopping online, you usually have to do your own research or comb through customer reviews to get an answer.
This could change in the future on Amazon through the use of artificial intelligence (AI). The AI assistant called “Rufus” is intended to make shopping on the site easier. The AI takes on a similar role to personal advice in stationary retail – for example, it answers product-specific questions and takes personal preferences into account. The assistant has been being tested with selected users in the USA since this year.
What does the use of AI mean for online trading?
At first glance, the use of AI in online trading brings significant advantages. According to Stephan Tromp, deputy general manager of the German Retail Association (HDE), AI contributes to the optimization and individualization of online shopping. The possible uses of AI are “diverse”.
And there would also be advantages for trading companies: the technology is particularly important in logistics and transport processes as well as in product optimization. “In retail, AI can help to precisely control the global flow of goods and to better understand customer needs locally in stores and online,” explains Tromp.
What can AI do?
AI does one thing special: it recognizes patterns. Based on the data available to her, she can recognize and analyze individual click and shopping behavior, explains Jörg Funder, retail expert from Worms University. This means the AI can assign customers with similar click behavior to a group and display products that “show a high level of correspondence with the click behavior.”
In addition, AI can draw conclusions about the properties of a product from customer reviews and return rates and offer customers tailor-made products, taking their personal preferences into account, thus making online shopping easier. For example, if many customers complain in online reviews about a top with particularly short sleeves, the AI records this feedback and does not even show this product to users who prefer long sleeves.
Few Germans trading company use AI
The advantages of AI in online trading are clear – but it has so far hardly been used in many German retail companies. According to a survey by the digital association Bitkom, 56 percent of the retailers surveyed see AI as a competitive advantage; However, it has so far only been used by four percent of all companies surveyed.
Other studies also show a similar picture: The 2023 AI survey by the German Retail Association shows that around 68 percent of retail companies do not use AI and do not plan to use it.
A complicated process
There are several reasons for the reluctance. “The implementation of AI projects often fails in practice due to the apparent lack of use cases. The high costs are also a major hurdle. Establishing AI in retail is a time-consuming and cost-intensive process,” says Tromp from the German Retail Association.
Trading expert Funder cannot understand this. Retail companies that forego AI also forego opportunities to increase efficiency. “Companies that do not invest sufficiently in AI will suffer a strategic competitive disadvantage in the medium term and will not be able to work as economically as their competitors. They will then exit the market later.”
The use of AI poses risks for consumers
A personalized and tailored online offer can be particularly convenient for customers; but it should not be accepted without reservation. The Federal Association of Consumer Organizations (vzbv) urges caution. The use of AI could increase the power gap between companies and individual consumers, warns Miika Blinn, consultant for the digital and media team at the Federal Association of Consumer Organizations (vzbv).
AI can be used to “recognize and exploit the personal weaknesses of individuals by attempting to influence purchasing decisions through individual communication.” And customer reviews should also be viewed with caution – the vzbv warns against fraud: “We see the danger that fake product reviews will persuade consumers to make decisions that they would not have made otherwise,” says Blinn.
AI Act decisive for the future of AI
One thing is clear: AI is changing online trading. And in order for the responsible use of AI to be possible in the future, an EU-wide uniform, legally secure and innovation-friendly set of regulations is particularly important, says Tromp. The so-called AI Act is particularly important. The European Parliament has now finally voted on the rules for regulating artificial intelligence. This also defines AI systems that are prohibited: including the use of subliminal, manipulative or deceptive techniques that distort behavior and impair conscious decision-making.
According to Blinn from vzbz, there are still gaps in protection and workarounds for entrepreneurs to use manipulative AI practices. How and whether a ban protects consumers from manipulation by AI is questionable. “This raises the question of how the rules of digital consumer protection need to be tightened up in the future in order to close these gaps,” said Blinn.