Shoppers navigating Black Friday 2026 are increasingly leaning on artificial intelligence assistants to hunt down holiday bargains, but independent testing shows that chatbot deal-hunting remains a mixed bag of broken links and phantom discounts.
As the holiday shopping weekend kicks off, millions of consumers are turning to tools like ChatGPT, Google Gemini, Claude, and Microsoft Copilot to scan promotions and compare prices in seconds. Behind the scenes, these platforms query recent web data, synthesize e-commerce databases, and rank recommendations based on discount depth.
Figures indicate that while tech firms have raced to roll out shopping features—such as Amazon launching Alexa for Shopping for UK users in November and OpenAI launching a Shopping Research tool on ChatGPT days before last year’s Black Friday—user adoption varies sharply by region.
A YouGov survey of 19 countries found that the UK ranked second-last in the proportion of shoppers using AI for retail, with 36% of Black Friday shoppers stating they did not plan to use AI tools at all.
When put to the test on a practical holiday shopping task, popular AI models yielded erratic results. Testing across four free-tier chatbot versions to find a 55-inch television deal exposed significant gaps in real-time inventory tracking and price accuracy. OpenAI’s ChatGPT produced four options, starting with a Hisense TV on Chinese retailer Joybuy’s website priced at £499 down from £649. Yet a second recommendation for a 55-inch TCL-branded TV directed users to a sidebar displaying prices for a 50-inch version ranging from £379 to £814.

A third suggestion pointed to the same Hisense model at full price on John Lewis and Richer Sounds, which ChatGPT labeled as a “best safer purchase” due to longer guarantee periods.
Clicking the retailer links revealed the television was no longer listed online, while a fourth option featured an out-of-budget, unavailable TCL TV alongside a single “one OLED option to compare” for an £1100 LG model marked as an ad. Meanwhile, Anthropic’s Claude provided minimal specifications and relied on a small pool of online reviews before directing users to broken links.
Broader performance testing highlights both the strengths and vulnerabilities of algorithmic shopping assistants. Independent evaluations by consumer advocacy groups indicate that ChatGPT successfully identified 72% of advertised Black Friday discounts within a sample of 500 products, though it also generated three instances of non-existent deals stemming from outdated training data.
Google Gemini achieved an 85% accuracy rate on mainstream electronics by integrating directly with Google Shopping data, whereas Claude struggled with niche product categories.
A persistent limitation across all tested platforms is the inability to detect stealth price increases, where merchants inflate base prices before applying holiday discounts. Automated assistants cannot access invitation-only flash sales, loyalty-program promotions, or real-time inventory changes. Free-tier users rely on data current as of October 2026, while ChatGPT Plus subscribers gain access to real-time web browsing.
| Platform | Tested Performance | Primary Limitation |
|---|---|---|
| ChatGPT | 72% accuracy on 500 product samples | Outdated training data generated phantom deals and dead links |
| Google Gemini | 85% accuracy on mainstream electronics | Tied to underlying Google Shopping data freshness |
| Claude | Short summaries with minimal specifications | Lacks direct retailer API access |
Retail analysts recommend adopting a hybrid strategy rather than relying blindly on automated outputs. Shoppers can utilize AI assistants for initial product research and feature comparisons, but should manually verify final prices on merchant websites and cross-reference deals with established aggregators such as CamelCamelCamel or Slickdeals.
Experts advise asking targeted questions with precise product model numbers, requesting direct sources for every recommendation, and avoiding chatbots for time-sensitive lightning deals where inventory shifts rapidly.
Earlier coverage: Police and Experts Warn of AI-generated Storefronts Ahead of Black Friday