Google's AI-powered search engine, AI Overviews, is facing scrutiny for spreading misinformation through its AI-generated summaries feature called Gemini.

Recent errors include harmful suggestions such as consuming rocks and endorsing smoking during pregnancy, and falsely claiming Barack Obama was a Muslim president.

Google has acknowledged these inaccuracies but maintains that most summaries are accurate, with ongoing efforts to improve the system.

Experts have raised concerns about the accuracy and reliability of the AI system, noting instances where it has promoted debunked conspiracy theories.

The incident underscores the challenges of managing large language models and emphasizes the need for oversight and verification processes to maintain user trust.

Google aims to reach 1 billion users by the end of 2024, but the current issues signal a step back in user experience as users are now advised to double-check AI-generated summaries using traditional search methods.

Competitors are closely monitoring the situation as they continue to vie for dominance in the AI race.