New Study Finds Average Accuracy of AI Search Tools is Only 60%

Tow Digital News Centre has recently studied eight AI search engines, including ChatGPT Search, Perplexity, Perplexity Pro, Gemini, DeepSeek Search, Grok-2 Search, Grok-3 Search and Copilot. They tested the accuracy of each tool and recorded the frequency of its refusal to answer. Researchers randomly selected 200 news articles from 20 news publishers (10 each). They ensure that each article returns the first three results in the Google search when using excerpts. They then perform the same queries in each artificial intelligence search tool and rate accuracy according to whether the search correctly quoted A) articles, B) news agencies and C)URL. The researchers then label each search on the basis of the accuracy from “fully correct” to “totally incorrect”. As can be seen from the figure below, with the exception of two versions of Perplexity, the performance of artificial intelligence is not satisfactory. Overall, the artificial intelligence search engine has 60% inaccuracies. (Source: cnbeta)