Below the Moderation Line: How Payment Networks and Infrastructure Providers Set Speech Rules
Photo: Esquilo, CC BY-SA 3.0, via Wikimedia Commons
Most debates about online speech focus on the platform: the post, the account, the community-guidelines notice, the appeal button. But a publisher can be silenced without any of those things happening. A payment processor can drop a merchant. A card network can tighten its rules. An app store can reject an update. A cloud host can suspend an account. A domain registrar can disable a name. Each decision may be described as a business or compliance matter, yet each can remove a publication from public view.
This is the infrastructure layer of speech regulation. It is quieter than platform moderation and often less visible to the people affected. It also deserves more attention from anyone concerned about digital rights.
The visible layer and the invisible one
Platforms are the front door of online expression. Infrastructure providers are the plumbing. Users see the front door. Publishers discover the plumbing when it breaks.
The key feature of infrastructure decisions is that they are usually framed as contract enforcement rather than content moderation. A payment processor cites its acceptable-use policy. A host cites its terms of service. An app store cites a developer agreement. These are private contracts, and in the United States private companies generally have broad latitude to choose whom they serve. But when a small number of providers control essential access, a contract term can function like a content rule.
Card networks as speech gatekeepers
Card networks sit at the center of online commerce. Visa and Mastercard do not typically decide individual publisher cases, but their rules shape what payment processors will accept. Processors, in turn, decide which merchants can accept cards. For publishers in adult content, subscription publishing, political fundraising, and other contested categories, the practical question is often not whether a platform will host the speech but whether a payment rail will carry the money.
Adult content is the clearest example. Many processors impose stricter documentation, higher reserves, or outright prohibitions. The result is that lawful publishers can be pushed toward high-risk processors with limited leverage. Political fundraising raises a different version of the same issue: platforms and processors may apply heightened review after controversy, and the review can freeze legitimate activity while the publisher tries to prove compliance.
Ad exchanges and the economics of acceptable content
Advertising systems add another layer. Ad exchanges and brand-safety vendors classify publishers and content. A news site covering controversial subjects may be labeled unsuitable for certain advertisers, which reduces revenue without any formal moderation finding. The publisher remains online, but the economic base narrows.
This is not always censorship in the ordinary sense. Advertisers have a right to choose where their messages appear. But when a handful of intermediaries set broad categories for 'unsafe' content, those categories become de facto editorial standards for the open web. A publisher may never see the classification, and appeals may be handled through business channels rather than public policy processes.
App stores, cloud hosts, and domains
App stores control distribution on mobile devices. A rejected app can lose access to a large share of its audience, and the rejection may rest on a broad rule about user-generated content, payments, or 'objectionable' material. Cloud hosts control the servers that keep a site online. Domain registrars control the address. Each can act on an acceptable-use complaint, a legal demand, or an internal risk assessment.
The striking thing about these layers is their interdependence. A site may survive a platform suspension and still lose its payment processor. It may survive a payment dispute and still lose its host. The more layers that fail, the less practical recourse the publisher has.
Three categories under pressure
Adult content is the most established case. Lawful publishers face age-verification rules, payment restrictions, and hosting policies that treat sexual content as inherently high-risk. The effect is not a single ban but a series of filters that make lawful operation expensive and precarious.
Political fundraising is another. Campaigns and advocacy groups depend on payment processors and platforms that may apply enhanced review during contentious periods. A freeze or delay can be framed as fraud prevention, but it can also suppress fundraising at a critical moment.
Controversial publishing is the broadest category. Independent outlets covering war, policing, immigration, or extremism may find that ad networks, payment providers, or infrastructure vendors regard their subject matter as brand-unsafe. The publisher is not accused of illegality. It is simply classified as a risk.
What recourse exists
For publishers, recourse is uneven. Contract claims are possible but expensive. Public pressure can work, especially for well-known outlets. Some providers offer appeals, but the process may be confidential and the standards vague. Regulators can intervene in limited circumstances, particularly where competition or consumer-protection law applies, but infrastructure providers are not generally treated as common carriers.
That leaves a gap. A platform moderation decision at least produces a notice that can be compared against public rules. An infrastructure decision may produce only a termination email and a contract clause. The affected speaker may not know which vendor made the call, what evidence was used, or whether a human reviewed it.
Questions for policymakers
The infrastructure layer is not immune to public scrutiny. Lawmakers who care about online expression should ask whether essential payment, hosting, and distribution services are being withdrawn on the basis of vague risk categories; whether affected publishers receive a specific reason and a meaningful appeal; whether concentration in these markets gives a handful of vendors the power to set speech rules for the whole internet; and whether transparency requirements should extend beyond social platforms to the companies that keep them running.
None of these questions has an easy answer. Private firms need room to manage fraud, illegal content, and financial risk. But a system in which speech can be switched off several layers below the public debate is a system that needs more daylight.