> it would of course be cheaper for me to simply scrape the website I need instead of paying one of the apis.
That's the business model of those APIs; for a segment of their customers it's cheaper to outsource the scraping to them than do it in-house.
As far as legality goes, it's in a bit of a grey area. What you do with the data is also a major factor, regardless of how the data was obtained, and I would argue this is more important than the "how" - forget republishing stories verbatim on your website, regardless of whether you're getting them via API, scraping, or manually copy/pasting. Other usages might be more permissible, and some might be totally fine for example. Keep in mind that the API provider doesn't have to be legal either - they could be anonymous or based somewhere out of reach of the US legal system and thus couldn't care less about any potential legal consequences.
