Unemployment Data Has Yet to Show AI Hitting New Graduates Hard

Ars Technica’s headline presents an expectation that AI would hit new graduates hard. [1] The same headline says unemployment data has so far pointed in the opposite direction. [1] The headline’s stated contrast is between the expected harm to new graduates and unemployment data that says otherwise. [1] Its wording is explicitly provisional: the conclusion is limited by the phrase “so far.” [1]

The source supports a restrained reading: current unemployment data does not match the expectation that AI would already hit new graduates hard. That is neither proof that AI has no labor effect nor evidence that graduates face no employment pressure; it only narrows what can be claimed from the reported data. The phrase “so far” makes continued measurement more appropriate than a definitive verdict. What the sources don't address: which graduate occupations were measured, what period the data covers, or how AI’s effect was separated from other causes.