A single automated request that reads a public web page can be lawful under one body of law and forbidden under another, because web scraping is governed by three doctrines that were written in different eras and never reconciled: the Computer Fraud and Abuse Act, contract, and copyright. This article explains each in plain language for the two audiences that need it—data collectors who want to gather information without ruinous liability, and website operators who want protections that actually hold. It traces the hiQ v. LinkedIn saga and its paradoxical ending, the Supreme Court's narrowing of the CFAA in Van Buren, the Bright Data contract rulings, the fact-versus-expression line drawn by Feist, and the wave of AI-training fair-use fights including Thomson Reuters v. ROSS, Kadrey v. Meta, and Bartz v. Anthropic. It maps the hierarchy of technical barriers from robots.txt to DMCA Section 1201, the common-law backstops of trespass to chattels and hot news, and closes with a practical risk spectrum for both sides and a detailed FAQ.