During recent tests, AI assistant ChatGPT failed to accurately recommend products featured on WIRED’s gear review pages. The bot suggested options that were not even mentioned in the official guides, such as recommending an LG QNED Evo Mini-LED TV as the best overall pick when it was not included in the guide at all.

WIRED’s Gear Reviews team takes extensive time to test products, ensuring readers get up-to-date and reliable information. ChatGPT’s errors could mislead users, leading them to purchase items that have not been tested by WIRED’s reviewers. This raises concerns about AI tools’ ability to provide accurate recommendations.

During tests, the bot recommended Apple AirPods Max 2 as WIRED’s pick for wireless headphones despite the fact that the company had not yet reviewed them. Similarly, when asked about the best laptops to buy right now, ChatGPT incorrectly suggested an older model of the MacBook Air as the top choice.

The accuracy issues highlighted in these tests underscore the limitations of current AI technology. While chatbots can provide quick answers and assist with general queries, they may not be reliable for complex or nuanced information like tech reviews where human expertise is crucial.