At the 7th International Conference on Freedom of Expression in Asia held 24-26 August 2022, Open Net co-hosted with Google a seminar on “Rights-Friendly Content Moderation in APAC” where the speakers spoke as follows:
Damar Juniarto from Indonesia spoke about the harms of MR5. He talked about 5 platforms that were blocked for failure to register with the authorities per MR5 and also the regulation fashioned after NetzDG which requires platforms to take down awfully vaguely defined “prohibited content” within 1 hour or 4 hours.
Arthit Suriyawonkol from Thailand talked about intermediaries’ duty to find out identity of users behind VPN. Also, personal data protection law’s legislation is being postponed. Finally, the requirement of record-keeping on how personal data has been processed has been given too many exemptions.
_________ from Cambodia discussed the government’s plan to build National Internet Gateway and its possible harms on freedom of speech.
K.S. Park focused on the role of the internet and the overarching principles protecting it, which would present guidance on how to evaluate the national initiatives in other countries as follows:
The internet’s civilizational significance has been confirmed by the Korean Constitutional Court which said that ‘the internet allows people to overcome hierarchy in offline in gender, social status, age, etc., and therefore makes democracy more representative.’ This means that the internet helps with substantive democracy going beyond formal one.
In order to protect this civilizational significance, powerless individuals should be guaranteed the same tools of mass communication as big businesses and governments so that they can overcome hierarchy. For that to happen, intermediaries should not be held liable for contents that they do not know about otherwise they will try to pre- or post-filter all contents uploaded or shut down the platform. (Normally, general torts/criminal accessory liability will already require knowledge but the framers of intermediary liability wanted to make sure that the bar is not lowered for internet statutorily or judicially.) If contents remaining are only the ones approved by intermediaries like the news stories are pre-approved by the editors in the legacy media, powerless individuals’ mass communication capacity is gone. To avoid this, EU passed the intermediary liability safe harbor in e-commerce directive of 2000. Also, the directive stipulated against loss of freedom of mass communication when platforms are put under “general monitoring obligation”, which would again make sure only the contents approved by intermediaries remain online.
Now, the NetzDG-type mandatory notice and takedown developing in Southeast Asian jurisdictions does not violate the text of intermediary liability safe harbor since the liability is imposed only for contents that the platforms are given notice of. However, platforms will err on the side of deleting the contents when notices are given because their mistaken non-action immediately means liability. This will again take down many lawful contents according to the subjective whims of the platforms, again depriving people of the tools of mass communication. In order to comply with the spirit of intermediary liability safe harbor, the liability should be imposed only on KNOWING maintenance of unlawful contents. Indeed, NetzDG-type laws impose liability for failure to take down in strict timelines instead of allowing the platform operators to fully develop knowledge of the legality or illegality of the postings. Asian countries should not legislate the NetzDG-type mandatory notice-and-takedown laws. Failure to take down contents that platforms know to be illegal can be remedied by operation of the baseline torts law or general criminal law.
Another worrying trend is administrative censorship. The UN Human Rights Council have repeatedly passed resolution that “what is protected offline should be also protected online.” This also means that we should not have speech police online just as we do not have speech police for offline activities. Offline police move in after the speeches have done harms but most administrative censorship is geared toward prohibiting speech from reaching the intended audience. Most often the administrative body have conflicts of interest in evaluating the legality of postings, and usually do not have capacities to make those decisions. Also, given the censorial purpose of administrative censorship, even the standard of illegality is set in very loose terms. For instance, MR5 uses “prohibited content” while Korean law on administrative censorship uses the standard “what is necessary for nurturing sound communication ethics”.
Other aspects to be careful about is protection of anonymity. Free speech will be chilled if privacy is not protected. Most important speeches are the ones made public. So content of speech has not privacy implications. What needs be protected to support free speech is anonymity. In Asia, internet real name laws are being attempted (e.g., 2002 Philippines). Also, Indonesia’s MR5 specifies warrantless unmasking of speech makers which again reduces anonymity. In Korea, we filed constitutional challenges both on registration of posters (internet real name law) and warrantless access to subscriber data, We would like to share our experience and resources.