The government has repeatedly argued that ISP policies that permit companies to examine the contents of e-mails, search for spam, or block prohibited activity are not "searches" under the Fourth Amendment. Thus, everybody’s communications could be scanned for prohibited information.

Of course, such a technology is ripe for abuse. While the Australian filtering scheme only seeks to identify and filter child pornography and terrorist information, there is no reason it could not — or should not — be deployed to also seek out communications related to, for example, drug trafficking, murder, or the unlawful transfer of copyrighted materials. After that, we can look for people distributing unlawful manifestos which promote views unpopular with or prohibited by the government. All are potentially criminal activities.

There are two main problems with the proposals by the Australian government — and the actions of the N.Y. Attorney General — despite their good intentions. First, the proposals essentially enlist ISPs to act as government agents to do something that the government itself would not lawfully be permitted to do. Second, it mandates — or, in the case of New York, merely encourages under threat of prosecution or embarrassment — the intrusion into the privacy of hundreds of millions of innocent users and hundreds of billions of innocent communications.

A modest proposal

Let’s face it, the ISPs want to do the right thing. They want to make the Internet as open a place as possible and not be legally responsible for how people use it, but also want to keep it reasonably safe. Nobody except the shadiest of operators would want to become known as a haven for child pornography. The same may be said, to a lesser extent, of other kinds of prohibited conduct.

So, rather than mandate, why not offer their subscribers a blocking service? Indeed, Section 230 of the Communications Decency Act already requires ISPs to notify their customers about the availability of blocking software. The ISPs can create "child-friendly" services that block all sex and violence — and presumably tweakable by the subscriber — in addition to a "responsible adult" version that blocks child pornography and other kinds of illicit conduct. Other ISPs could choose not to, and offer completely free and open access to the Internet — kiddie smut included.

By making filtering optional, the ISPs would not be any more liable for the content that their subscribers download and would put the power to block content in the hands of the subscriber. This would limit the problem of inadvertently or accidentally finding prohibited content. Since the vast majority of subscribers would presumably want to block this content, eventually the marketplace would likely make the illegal content difficult to find.

However, in order to implement such filtering, some tool would still have to examine the contents of every communication. ISP privacy policies would have to make it abundantly clear that this tool was examining everything the user was doing during the process of filtering content. It would also be recommended for the agreement to also state that the consent to such automated monitoring did not constitute a waiver of any rights to privacy in the contents of communications. Thus, for the government to get the contents of an email, a bittorrent or an IM, the normal legal processes would still apply.

The end result is that we would have enlisted, as Adam Smith called it, the invisible hand of capitalism, rather than the paternalistic fist of the government.