A recent series of incorrect answers provided by Google’s new “AI Overview” tool has sparked concerns among experts, raising alarms about potential dangers, particularly in critical situations such as health emergencies.
Previously, when users queried Google about obscure topics, the search engine would provide a ranked list of websites, allowing users to explore and verify the information themselves.
However, with the introduction of the AI Overview feature, Google now generates instant answers using artificial intelligence, leading to instances of misinformation and inaccuracies.
For example, when a Press reporter asked Google if cats have been on the moon, the AI-generated response claimed that astronauts had indeed encountered cats during moon missions, a statement entirely devoid of truth. Similarly erroneous responses have been shared on social media since the introduction of AI Overview.
These inaccuracies have raised concerns among experts, who fear that the new feature could perpetuate bias, spread misinformation, and potentially endanger individuals seeking help, especially in emergency situations.
In one instance, when AI researcher Melanie Mitchell queried Google about the number of Muslim presidents in the United States, the response confidently asserted a debunked conspiracy theory, attributing one presidency to Barack Obama. Despite attempts to verify the claim through a cited academic source, it was evident that the AI system failed to discern the credibility of the information.
Critics argue that Google’s AI system lacks the sophistication to differentiate between reliable and unreliable sources, leading to irresponsible dissemination of misinformation. While Google has assured swift action to address errors violating its content policies, concerns persist regarding the broader implications of AI-generated responses.
Despite claims by Google that the AI Overview feature has undergone extensive testing and provides high-quality information in the majority of cases, experts caution against relying solely on AI for critical information, particularly in time-sensitive situations.
Moreover, there are broader concerns about the impact of relying on AI for information retrieval, including the potential erosion of human curiosity, literacy about online content, and the disruption of online forums reliant on Google traffic.
As Google faces scrutiny over the accuracy and reliability of its AI-powered search features, competitors such as OpenAI and Perplexity AI are closely monitoring the situation. Perplexity AI, in particular, aims to challenge Google’s dominance with its own AI question-and-answer app, highlighting the growing competition in the AI-driven search landscape.
The incidents surrounding Google’s AI Overview feature underscore the need for rigorous oversight and accountability in the deployment of artificial intelligence in critical information services.