Page 233 - A Study on the Role of UGC Platforms in Copyright Law:An Intermediary-oriented Approach
P. 233

A Study on the Role of UGC Platforms in Copyright Law:   Chapter 7 Platform Users’ Entitlement to UGCs: Human Use and Web Scraping
 An Intermediary-oriented Approach

                 members. Neither would it be easy to distinguish users’ family members from web scrapers.
                 More importantly, the essence of code-based authorisation rests in the notion that anyone,
                 either a human or a scraper, can access information with a code. As web scraping is a faster
                 version of human browsing with fewer mistakes, but has no qualitative differences from
                 human browsing, a human browser provided with a code can access the same information
                 to the same extent as a web scraper provided with a code. If a human’s access to a system
                 with a shared password is considered authorised, so is a web scraper’s access. Therefore, I
                 endorse Judge Reinhardt’s dissenting opinion in Nosal II:


                    While the majority reads the statute to criminalise access by those
                    without ‘permission conferred by’ the system owner, it is also proper
                    (and in fact preferable) to read the text to criminalise access only by
                    those without ‘’permission conferred by’ either a legitimate account
                    holder or the system owner. 155

                    (4) Incompatibility of the CFAA and web scraping
                    The above account explains the fierce controversies that have arisen when the CFAA has
                 been applied to web scraping, which has made claims filed under the CFAA highly uncertain
                 and has even resulted in unreasonable decisions. Despite the efforts to restrict the scope of
                 the CFAA, using CFAA as an ‘expansive misappropriation statute’ still sends the wrong
                 signal that the law intends to ‘criminalise large groups of people who would have little
                                                                156
                 reason to suspect they are committing a federal crime’.  The debate has grown out of the
                 impossibility and unreasonableness of applying an outdated, criminal anti-‘hacking’ statute
                 to commercial web scraping practices that are commonplace and even desirable in the big
                 data era.
                        157
                    The CFAA has not only broadened the reach of criminal liability; it also has the potential
                 to override freedom of speech. Citing the US Supreme Court’s decision in Packingham v.
                 North Carolina, holding that ‘to foreclose access to social media altogether is to prevent
                                                                                          158
                 the user from engaging in the legitimate exercise of First Amendment rights’,  the
                 American Civil Liberties Union (ACLU) asserted that the access provision in the CFAA
                 is constitutionally overbroad and criminalises acts protected by the First Amendment.
                                                                                             159
                 A concern more related to the UGC database context is that the CFAA is likely to exert
                 prohibitive effects on innovation by encouraging data monopolies. In a world of algorithms
                 and artificial intelligence used for big data analysis, criminalising access to data connotes a


                 155  United States v. Nosal, 844 F.3d 1024, 1052 (9th Cir. 2016).
                 156  United States v. Nosal, 676 F.3d 854, 859 (9th Cir. 2012).
                 157  Williams, ‘Automation Is Not “Hacking”’ (n 118) 428.
                 158  Packingham v. North Carolina, 137 S.Ct. 1730, 1737 (2017).
                 159  Sandvig v. Sessions, 315 F. Supp. 3d 1, 13 (D.D.C. 2018).


                                                                                          • 219 •
   228   229   230   231   232   233   234   235   236   237   238