It's actually that I "read" the data, regardless of my intention and regardless of what I intend to do with it. But taken to the extreme there has to be something that makes it legal for me as a consumer to scrape ONE item from their data, while at the same time making it illegal for a company to scrape ALL their data in this way.
The worst situation to be in is one where it's "legal until it isn't" i.e., when you actually cause a problem (bandwidth, revenue) you will be sued - and until then you no idea.
Technically (and it's complicated these days because copyright law was mostly written well before computers and the internet existed) any time you create a copy of someone else's creative work, you need an explicitly granted right to do so. You can _read_ a book, but you cannot write down all or part of what you just read. You can listen to a song, but you cannot perform all or part of that song (An Australian band called Men At Work lost a copyright case for a flute solo that contained 2 bars/11 notes of a 1930's folk tune "Kookaburra" who's copyright had been sold/bought in 2000.)
There are some explicit legal exceptions to requiring a license, most commonly people talk about "fair use", but that's again a "pre internet" body of law, and is extremely open to interpretation as to how it applies to digital copies in computers.
https://www.copyright.gov/fair-use/more-info.html
https://fairuse.stanford.edu/overview/fair-use/what-is-fair-...
Example: Does it matter whether the data is ever stored? Say I make a listing notifier that pings people when a house is listed in their desired area. I’d need to crawl real estate listings and but I’d never store or republish any content from the site.