An illustration showing a person using a laptop with artificial intelligence chatbots displaying retail products and price tags.
Technology

Can AI really help spot Black Friday deals? We put popular models to the test

Illustration: NexVani Media

By: NexVani DeskDate: 21 November 2026Views: 0

We put four generative AI models to the test to see if they could successfully help us find a 55-inch television bargain ahead of Black Friday.

Finding the best bargain during the annual retail event can often feel overwhelming, with shoppers wading through endless online choices, fake reviews, and questionable discounts. Generative artificial intelligence products have increasingly promised to assist consumers in navigating these online shopping woes. A week ahead of the retail event, we decided to test several free models to see if they could actually help us track down a 55-inch television deal.

Despite the growing interest in technological tools, current figures indicate that relatively few shoppers utilize them for this purpose. A survey conducted by YouGov across 19 countries revealed that the UK ranked second-last in the proportion of people turning to artificial intelligence for shopping help. Among respondents shopping around Black Friday, 36 percent did not plan to use any such tool at all. Nonetheless, consumer interest is rising, leading tech firms to introduce dedicated features. Amazon launched Alexa for Shopping for UK users, OpenAI released a shopping research tool on ChatGPT, and Google introduced Universal Cart in the US with price tracking and promo code features.

Generative artificial intelligence models produced mixed results when asked to find a television bargain.

Testing the free versions of popular chatbots yielded inconsistent and often frustrating results. When asked to find a 55-inch TV deal, ChatGPT listed a Hisense television from a Chinese retailer at a discounted price, followed by a TCL television with conflicting sidebar pricing. Another suggestion was the same Hisense model at full price from traditional electronics sellers, which ChatGPT favored for longer guarantees, though the links led to pages stating the product was no longer listed. Another recommendation was an unavailable TCL model exceeding our budget, while a final suggestion was an advertised OLED option priced at ยฃ1,100.

Claude provided the briefest response without detailing technical specifications, seemingly relying on a small pool of online reviews. Clicking its links led to dead pages on a retailer website and year-old promotional round-ups. Claude also expressed hesitation regarding two recommendations due to a scarcity of independent reviews. Critics have noted that verifying chatbot information can consume extra time, while brands actively optimize their presence to appear in generated responses.

Why it matters

This experiment highlights the current limitations of relying on generative chatbots for consumer purchases. As major technology companies roll out specialized shopping features, shoppers turning to these tools may encounter dead links, outdated information, and unavailable stock rather than guaranteed savings.

Add NexVani Media as a preferred source on Google

Sources

ND

NexVani Desk

NexVani Media brings you the most important world news as it happens, with every story linked to its sources.