AI hallucinations are not a new word, but somehow, you will always find Google AI connected to it. Last year, the AI ​​Tool made headlines, and not for good. It was infamous to suggest users to glue the pizza. And now, the phase is back. Social-media platforms are filled with screenshots, where Google AI cannot understand which year it is. While we are on half the path through the year 2025, Google AI overview is still living in 2024.

At the time of writing this article, Google had already solved the issue. But, to what extent? To test the Google AI overview, we asked the tool, “Is this Saturday?” On 31 May, 2024. And this is what he said.

While it has identified the place correctly, time and date is still wrong. On X (East was known as Twitter), users are posting similar questions, the most popular, “Is this 2025?”

Advertisement

While Google announced the year as 2024 for some users, some users posted confused answers in the AI ​​overview. A user, @Lillyraynyc posted a screenshot of the same querry and for that, Google replied differently. “Is this 2025?” When asked, it pops up with the answer, “No, it’s not 2025. The current year is 2025.” I mean, take a break, AI overview.

Sometimes wrong and sometimes misleading answers by an AI tool definitely increase eyebrows, given that Google is working hard to bring the “A” game in this race. Techcrunch was the first publication to inspect the trend. According to the report, a spokesperson of Google told Techcrunch, “With all the search features, we strictly improve and use such examples to update our system. The vast majority of AI overview provide auxiliary information, and we are actively working on an update to solve this type of issue.”

When asked about the cause, Google did not offer a specific explanation, simply saying that it is working on an update to prevent similar problems in the future.

As stated earlier, this is not just a case where Google AI overview has talked. Among their more colored blunders, AI suggests that users eaten “a small rock per day” to get their dosage of minerals, and even recommended to use glue to keep cheese from slipping a pizza. While these moments have given birth to a lot of memes, they also highlight the very human side of machine intelligence.

Nevertheless, despite these AI hiccups, CEO Sundar Pichai says the equipment is being completed. He recently revealed that the AI ​​overview has now reached 1.5 billion users in more than 100 countries. In major markets like the US and India, it is responsible for more than 10 percent of the relevant discovery questions.

He said, this example acts as a clear reminder that AI is far from innocent. At a time when major technical companies like Microsoft, Google and Meta are allegedly cutting jobs and rapidly turning to AI-operated solutions, these misunderstandings provide a small feeling of confidence- the human decision still matters.