Google AI has a very high rate of factual errors that are easily verified with public data. For example it repeatedly puts a Georgia city in the wrong county such as Griffin in Carroll County.
Given its track record for factual data, it is not worth trusting for opinions, evaluations. A large number of IT workers are not concerned about GIGO. That trait seems to be passed on to AI.
The code examples generated by AI on Edge should be regarded as suggestions not as authoritative. I suspect that would be the same as on Google.