The article made me question Google search even more.
« we tested the feature extensively before launch. This included robust red-teaming efforts, evaluations with samples of typical user queries and tests on a proportion of search traffic to see how it performed »
Then she explains that:
«We’ve also seen nonsensical new searches, seemingly aimed at producing erroneous results.»
So, the red-teaming didn't think of prompt injections or how to identify fake results? If that's the case, wouldn't that mean all results from Google searches are questionable, considering how much crap coming out of SEO-farming websites?