While many hope that Web3 technologies will transform the internet by decentralizing and redistributing control of essential services and infrastructures, the problems of moderating user generated content aren’t going anywhere. Content moderation has been a major concern for Web2 platforms and will almost certainly be one for Web3 apps as well. Those running NFT marketplaces, decentralized social media apps, file storage nodes, and anything else that involves user generated content will have to make decisions about which content, users, and behaviors to permit and which to remove.
Examining Web3 content moderation, this blog post is the first in a series exploring Web3 Trust & Safety issues. Content moderation presents a series of difficult questions and problems. Rather than offering a comprehensive or definitive take on the matter, this post opens the conversation with a 40,000-foot view of Web3 content moderation, presenting a few thoughts on what might be new about Web3 content moderation and what is not.
1. Web3 is and will likely be a diverse set of apps, communities, and technologies; there will be no one-size-fits-all solutions to content moderation.
Amid the explosion of interest in Web3 over the past year, discussion often treats Web3 as though it is a singular technology. But Web3 is still in the process of being imagined, designed, and built. Much of Web3 remains hypothetical.
But also, whatever it will be, it will not be singular. If the core goal is to build a more decentralized internet, almost by definition this means there will likely be many different types of apps, infrastructures, and communities that embrace different technological designs, norms, structures, and governance systems. As concerns content moderation, there is both risk and opportunity in this diversity: while there are unlikely to be any one-size-fits-all solutions, there will be plenty of space for creative experimentation with new moderation approaches.
2. Web3 apps will face significant pressure from countries, partners, and users to have detailed content policies and enforcement systems.
Broadly, there are three types of pressures on Web2 companies to develop detailed content policies and moderation practices. These pressures will also likely encourage Web3 developers to develop content moderation plans and tools.
First, many countries have laws prohibiting certain types of speech. Platforms must be able to remove this illegal material. For example, while the US government imposes few restrictions on speech, some things, including child pornography, terrorism, and fraud are prohibited. Section 230 of the Communication Decency Act cannot be used as a defense for intermediary liability regarding content that violates federal criminal law. Intermediaries must answer for the illegal content they knowingly leave up.
Second, many companies encourage or require partners to restrict and moderate content. Recently, AWS and the Apple and Google app stores all banned Parler for deficiencies in its content moderation efforts. These actions are occasionally referenced by Web3 proponents as examples of the too-powerful companies Web3 will ultimately replace. But while decentralization might mean there are more companies, communities, and infrastructure providers, it won’t change the desire of (some) to only work with partners that prevent illegal and illicit material.
Third, many of us would rather be in online communities free of illegal or problematic content, spam, fraud, and harassment. In being based on open protocols, Web3 technology may allow users more ownership and control over their data and greater ability to switch between apps that provide the services and environments they want.
3. There are already a range of approaches to Web3 content moderation.
Some Web3 apps and communities have already begun crafting content policies and enforcement systems. For example, Coinbase recently released a statement on their account removal and content moderation policy (see figure below) that offers a minimal approach rooted in removing illegal content and activity.
