Should Internet Platforms Be Classified as Common Carriers?

In my latest blog post, I illustrated that the increased privacy rationale used by the majority of the commissioners at the Federal Communications Commission to support the reclassification of broadband internet service providers (ISPs) in its resuscitated Open Internet Order, rulemaking would have negligible effect. This is because the risk of privacy breaches by ISPs (for example, selling private information and location data) is small compared to the risk faced from the providers of applications hosted on the consumers’ internet-capable devices (computers, tablets and smartphones) and the holders of their information on corporate and cloud-based servers. This begs the question, If ISPs must become subject to Title II of the Communications Act of 1934, then why not also the internet platforms used by the vast majority of consumers as their primary interface with the internet world?

Title II of the Communications Act of 1934 requires stricter obligations on certain providers deemed to be “common carriers.” The concept of common carriers has typically applied to entities such as telephone companies, broadcast stations, and wireline networks. They are required to offer their services to all customers without discrimination or censorship, because they are essential services that should be available equally and without restraint to all who seek to use them.

Arguably, many internet platforms, such as social media sites, massive international e-commerce providers, and information search and access tools, are equally essential to the full participation of citizens in a digital society—both as consumers and producers. For example, when Facebook removed all links to Australian news content from its platform on February 18, 2021, the collateral damage included the removal of pages for charities, small businesses, public services, government departments, and even a range of emergency services. If the FCC’s proposed rulemaking is going to successfully safeguard citizens from internet outages by requiring ISPs to become subject to Title II provisions, then arguably it must require monopoly and essential service internet application providers to adhere to the same standards.

So, what might be the outcome of making significant internet platform providers subject to common carriage provisions? These include:

Non-discrimination: To provide their services to all customers without unjust discrimination. They must treat all customers equally and cannot refuse service based on factors like race, gender, or religion.
Access: Provide access to their services to the public without any unnecessary restrictions. They cannot refuse service based on arbitrary or unreasonable criteria.
Pricing: Common carriers are typically regulated in terms of their pricing practices. Regulators may set guidelines or investigate pricing practices to ensure they are fair, non-discriminatory, and not exorbitant.
Service Quality: Regulators may impose performance standards to ensure that common carriers provide a certain level of service quality to their customers. This can include requirements for reliability, availability, and responsiveness.
Safety: Common carriers are often subject to safety regulations to protect users and the public. These regulations may include maintenance standards, training requirements, and safety protocols.

Clearly, nondiscrimination and access provisions would constrain the ability of platforms to censor content on their sites. All content would have to be carried, so long as it met legal requirements. Just like the postal service cannot refuse to carry publications promoting distasteful, but still legal, content, neither could the platforms refuse to post it. It would be impossible to unreasonably deny access to the platform to certain individuals. The criteria under which reasonable denials could be achieved would have to be transparent, subject to regulatory approval and applied without any discrimination. Adherence would be monitored by the regulator.

If content is to be treated in a non-discriminatory manner, then it begs the question of whether internet platforms could continue to engage in prioritization of one piece of content over another—whether such prioritization is paid for or not. This includes the explicit de-prioritization of distasteful, but otherwise legal, content. It would also appear to undermine the rationale for paid advertising on internet platforms, as this requires the prioritization advertising content pushed by the platform to specific users. The efficacy of the model is underpinned by discrimination amongst both recipients and advertisers. Again, transparency would be required to demonstrate that the fees charged (or rewards paid) to advertisers and subscribers were “reasonable.”

If nothing else, this hypothetical exercise and the preceding blog post demonstrate that common carriage provisions developed in the 1930s are not well-suited for regulating either internet platforms or ISPs. Common carriage rules may be helpful for regulating the movement of physical goods, and may have had some use in a single application such as voice telephony. Applying them blindly may prove to be either pointless (in the case of privacy protections) or prone to inducing counterintuitive or unwanted outcomes, because the technologies concerned are far more complex and wide-reaching than those for which they were designed and intended.

Caveat Legem.