Leaked TISA safe harbor proposal: the right idea in the wrong place
Electronic Frontier Foundation | 31 January 2017
Leaked TISA safe harbor proposal: the right idea in the wrong place
by Jeremy Malcolm
A new leak of the Electronic Commerce chapter [PDF] of the Trade in Services Agreement from the November 2016 negotiating round has exposed a brand new U.S. government proposal on Internet intermediary safe harbors. The proposal, which the European Union is shown as opposing, is a rough analog to 47 U.S.C.§ 230, enacted as part of the Communications Decency Act (known simply as "Section 230", or sometimes as CDA 230).
Section 230 is one of the most important provisions of U.S. law for online platforms that host users’ speech. It provides a shield protecting online intermediaries against a range of laws that would otherwise that would otherwise hold them responsible for what their users say or do online. Although there are exceptions to this law—for example, the immunity does not protect platforms’ hosting of user-generated material that infringes copyright (which is governed by the weaker DMCA safe harbor)—Section 230 remains an invaluable catalyst to innovation and free expression online, and a major reason for the success of U.S. Internet platforms around the world.
The existence of a U.S. proposal for TISA based on Section 230 had been rumored for some months, and when asked directly about it last October the USTR did confirm its existence to EFF. However, we had not seen a copy of the text until now. Like Section 230, the provision excludes intellectual property rights and criminal law enforcement, but otherwise provides:
2. [N]o Party may adopt or maintain measures that treat a supplier or user of an interactive computer service as an information content provider in determining liability for harms related to information stored, processed, transmitted, distributed, or made available by the service, except to the extent the supplier or user has, in whole or in part, created, or developed the information.
3. No Party shall impose liability on a supplier or user of an interactive computer service on account of:
- any action voluntarily taken in good faith by the supplier or user to restrict access to or availability of material that is accessible or available through its supply or use of the interactive computer services and that the supplier or user consideres [sic] to be harmful or objectionable; or
- any action taken to enable or make available the technical means that enable an information content provider or other persons to restrict access to material that it considers to be harmful or objectionable.
Although we usually talk about Section 230 in the context of the protection that it provides platforms for hosting or republishing the speech of users (paragraph 2 above), it also does the reverse—protecting them from liability for removing users’ speech from their platforms, provided that they do so in good faith (paragraph 3 above). This so-called "Good Samaritan" provision affirms that online platforms are entitled to choose what user content they do or don’t wish to host, and allows technology providers to provide tools for platform owners to use in exercising that choice. Without this legal clarity, Internet intermediaries could face legal consequences for choosing not to host or provide access to content that they find objectionable on their platforms or networks.
EFF is a supporter of the Section 230 safe harbor, and we would also support its extension to the other TISA countries that presently lack similar protections for Internet intermediaries in their law. Just to give two examples from countries that are amongst TISA’s negotiating parties, Turkey frequently threatens Internet platforms such as Facebook and Twitter with liability for the speech of their users, and in Estonia an online news publication was held liable in defamation for anonymous comments submitted by users. Such claims against Internet platforms would fall flat in the United States, thanks to the Section 230 safe harbor.
But it’s for this reason, probably, that Europe is opposing the TISA proposal. Like the United States, Europe goes into trade negotiations with the express objective of maintaining its existing laws, and Europe’s equivalent to CDA 230, its E-Commerce Directive, simply doesn’t measure up to this U.S. proposal. Although Europe is also considering adopting a Good Samaritan provision to clarify that providers will not become liable for user content by reason of steps they take to filter out and eradicate illegal content on their platforms, there is no similar proposal to expand safe harbor protection for user content that intermediaries leave online. Indeed, if anything, Europe is planning to lump intermediaries with additional responsibility for user content.
It’s likely, then, that this proposal is either dead in the water, or else that it will be considerably watered down before TISA is finalised, if ever. And there in a nutshell lies the reason why EFF, despite our support for Section 230, can’t support the inclusion of this provision in a closed, secret trade agreement such as TISA. It is by pure good fortune that we have been able to read this first draft of the USTR’s proposal thanks to the document being leaked. But unless and until it is leaked again, we will remain blind to any changes that may be wrought in the back and forth of TISA’s closed-door negotiations, which might well end up twisting the proposal beyond recognition.
EFF commends the USTR for the intent of its proposal. We too have promoted the extension of Section 230-style safe harbor protection around the world, through our Manila Principles on Intermediary Liability. But until trade agreements can be made more open and inclusive, they are the wrong tool to promote such an important policy for the global Internet.