Last week, the Federal Communications Commission announced that it would soon release an NPRM on consumer privacy. According to the Commission’s “fact sheet,” the Notice will propose a number of requirements on broadband Internet Service Providers, including an obligation not to share customer information with non-affiliated services without first obtaining the customer’s express consent.

I may have more to say about the FCC’s proposed privacy rules when the NPRM is released. For the moment, however, I wanted to raise a broader issue that the proposed privacy rules, as well as AT&T’s swift response, put in focus: whether it makes sense to single out broadband ISPs for special regulation when other players in the Internet “ecosystem” engage in similar behavior. I predict that this will be one of the recurring debates in communications regulatory policy in the years to come.

Privacy provides a nice context in which to approach this question because it is already so well known that other Internet players, and not just ISPs, make frequent use of consumer information. Just think of search engines like Google, social media platforms like Facebook, and “sharing economy” services like Uber. But the Commission—for a mix of legal, historical, political, and policy reasons—isn’t prepared to regulate those services, and the fact sheet makes clear that it won’t. Thus, such “edge” providers will continue to be regulated by background consumer protection and antitrust law—administered by the FTC and DOJ—not by the FCC, and the substance of the two legal regimes may end up looking very different. The result is a rather fragmented looking regulatory regime once one pulls back the lens and views the system as a whole.

Differential regulation of ISPs isn’t only an issue in the privacy context. Take network neutrality, which has dominated the FCC’s docket as of late. The core concern of network neutrality proponents is that ISPs will use control over their networks to favor certain content or service providers over others. But companies other than ISPs might similarly serve as bottlenecks at different layers of the Internet, and have similar incentives to discriminate that ISPs have. Google, of course, most readily comes to mind, and in fact various companies have complained over the years that Google has discriminated against them in favor of Google-affiliated content. Some scholars have even raised the possibility of a “federal search commission” to deal with such issues.

The most common argument for differential regulation of ISPs has to do with fixed costs and, relatedly, competition. Though the United States often relies on competition to protect consumers, there are (typically) at most two providers of wired broadband in a given area, the duopoly market structure is hard to break because of the very high costs of laying wire, and even when there is more than one provider consumers often cannot or do not switch among them. For those reasons, we should be more skeptical that competition for consumers can fix market imperfections and should resort more quickly to industry-specific regulation when it comes to ISPs.

This argument is often a good one, and, at the very least, should make us pay closer scrutiny to the practices of ISPs compared to other players. We should also be careful not to simplify matters, however. Though we are not there yet, in the future it is possible that superfast wireless networks, municipal broadband networks, and new private options like Google Fiber may bring more competition to broadband markets, and we should make sure that regulation encourages rather than discourages those kinds of investment. And on the other side, it is often wrong to assume that “digital” services operate in a state of near-perfect competition. For one thing, network effects can serve to attract and lock-in customers to already-popular digital platforms, making market concentrations more durable. It’s also not right to think that today’s market winners have little more going for them than good code. Google and Facebook, for example, operate massive server farms in order to efficiently deliver their services around the globe, in some cases costing them upwards of $1 billion a pop.

One final note: The observation that we single out ISPs for behavior also exhibited by other Internet players is often used, at least rhetorically, as an argument against regulation at all—that is, as an argument that we should “level down” regulation. That doesn’t necessarily follow, however. One can eliminate differential regulation, after all, by leveling up as well as by leveling down. And, as indicated above, sometimes differential regulation will turn out to be justified. In any event, whether we should level down, level up, or maintain some amount of differential regulation in particular areas is likely to be a continuing debate in communications policy for the near future.