A good short take except the OUP's sales pitch at the end.

Why do they use AI?

Because they lack knowledge of the subject matter.

Why do they struggle to spot misinformation?

For the same reason.

The AI paradox --- Results are probabilistic and inherently unreliable. The only rational way to spot inevitable shortcomings is research and learn the subject matter --- so you don't need to rely on AI results.

If the results are really important, accepting what AI provides at face value is a recipe for catastrophe.