Today is Thursday, May 29, 2025. I hope we all can agree on that. But Google’s AI Overviews are seriously confused about what today is.
Is it Tuesday? Here’s what Google thinks:
Google thinks today is Wednesday and yesterday was Tuesday. But it got the dates correct (today is May 29 and yesterday was May 28). So, yay?
Google also spit out this equally helpful answer, telling me the date is May 29 (correct) but again said that it’s Wednesday, even including a citation to a site.
When I clicked on that link, it took me to Google’s homepage, which did not have any information about what day it is:
Wow.
Is it Wednesday? Is it Thursday? Apparently, these questions were too hard for AI Overviews to answer. Nothing triggers on these. (Also, no results for [is it Saturday]).
Is it Friday? Google gets it right here. Tomorrow is, in fact, Friday.
Is it Sunday? Monday? No AI Overview, but Google helpfully tells us that Sunday is June 1, 2025.
As for Monday, well, Google tells us today is May 29, 2025:
Is it 2025? More cringeworthiness, as shared by Lily Ray, vice president of SEO and strategy at Amsive, in a LinkedIn post – yes, that’s Google telling us that it both is and isn’t 2025:
Think it couldn’t get worse? Well: based on “a calendar,” Google AI Overviews tells us that the current year is 2024 and it is May 28 (as spotted by Lauren Donovan, director of marketing at Third Door Media, publisher of Search Engine Land):
But have no fear – Google gets it right if you ask [is it 2021?]. However, Google also gets it wrong because May 27 was two days ago.
Yikes.
Reaction. SEO consultant Andrew Cock-Starkey probably had one of my favorite takes on these awful AI-generated answers in a LinkedIn post:
“Google are so terrified of being left behind in ‘The AI Race’ (or at least being perceived to be left behind) that they’re literally building in public. On their biggest and most famous product.
And it’s cringeworthy.
Huge data centres, nuclear power, rinsing through fresh water, torching energy at a rate of knots… for this?”
Nikki Lam, head of SEO at NP Digital, also nailed it with this LinkedIn post:
Why we care. “AI responses may include mistakes.” No kidding. AI Overviews continue to disappoint, even on some basic queries that a 4-year-old human could fairly easily answer. Despite all the hype and regardless of how Google wants to spin the awesomeness of its AI models, AI Overviews are a flawed product that is contributing to the death of the current business model of the web.
