Opportunities and risks when shopping with AI.

This past week, many American news outlets have been reporting on a new issue in e-commerce related to AI: artificial intelligence can also be exploited to create fake sellers, advertisements, reviews, and even sophisticated scams.

"Can you help me find a compact camera, or a device like a gimbal at a good price for convenient shooting in crowded places?" Try searching for products with an AI virtual assistant; you can see that AI is almost like a personal shopping consultant, understanding user needs, helping compare options, and even suggesting the best time to buy.

Major American tech companies are seizing this growth opportunity. The tech news site The Verge mentions Google's "Global Shopping Cart" app. This shopping cart also tracks prices and provides notifications when items are in stock.

According to CBS News, Amazon developed the "Rufus" assistant to track product prices on its online retail platform, recommend prices based on demand, and complete transactions. Meanwhile, Walmart, the largest retailer in the US, also launched an AI assistant capable of conversation, providing advice, and evaluating products.

Business Insider highlighted that "the next large-scale AI war could be happening right in your shopping cart." AI platforms are projected to drive e-commerce sales in the US to $144 billion by 2029.

AI virtual assistants can help consumers find and ask for product advice, but they can also provide misleading information or even be exploited for online fraud.

CBS News posed the question, "What kinds of mistakes can happen?" For example, a company owner paid $30,000 to speak at the World Economic Forum in Davos, Switzerland, using an AI assistant. The AI assistant's founder explained that the user's instructions contained conflicting guidance.

The tech news site The Verge warned about a case on TikTok where a woman cried and pleaded with netizens to buy her products to save her business, but it turned out not to be a real person.

The Wall Street Journal cited a humorous image of an AI and a masked criminal as a pair. Last year, the FBI's Cybercrime Complaint Center received more than 22,000 complaints about AI-related scams.

AI virtual assistants should only be tools for support and reference, not for making purchasing decisions on behalf of buyers. At the very least, before clicking "buy," they should ask, "Is this information correct?" to ensure accuracy and to train the AI to become a safer version.

Source: https://vtv.vn/co-hoi-va-rui-ro-khi-mua-sam-bang-ai-100260602105126096.htm