AI Chatbots Deceiving Users with False Information: Google Report

Google’s head of search, Prabhakar Raghavan, has cautioned against relying solely on AI chatbots for providing accurate information. In an interview with the German publication Welt Am Sonntag, he stated that these AI-powered systems may deliver misleading and convincing answers.

Raghavan described this phenomenon as a “hallucination,” where the machine delivers a convincing but entirely false response.

However, a commercial for Bard was seen providing an incorrect response to a query regarding the James Webb Space Telescope. The mistake resulted in a significant decline in Alphabet’s stock, along with mounting concerns about the potential harm ChatGPT could bring to Google’s search dominance.

Raghavan told the newspaper, Welt Am Sonntag, that Google was acutely aware of the need to release Bard to the public, but also felt a great sense of responsibility. He said they didn’t want to misinform the public.

“We are exploring ways to integrate these possibilities into our search functions, especially for queries that don’t have a single definitive answer,” he added.

Raghavan’s statement follows criticism of Google and its CEO, Sundar Pichai, for the hasty and unsuccessful launch of Bard.

A former researcher at Alphabet’s AI division, Google Brain, Maarten Bosma, stated on Twitter that the demonstration indicated the company wasn’t treating AI with the necessary seriousness.

Insider attempted to contact Google for a comment, but the company did not respond due to being outside of normal business hours.

Leave a Comment