David Fidler has a new brief for the Council on Foreign Relations examining U.S. public and private sector efforts to counter the online propaganda of the Islamic State, a militant group also known as ISIS and ISIL. He finds that there is a lot of uncoordinated activity but not much evidence that it is doing any good.

“Radicalization usually involves more than consuming extremist tweets”

Many policymakers argue that online propaganda helps to radicalize young people. Fidler points out that there isn’t much strong evidence to support this claim. Internet-based communication is plausibly much less important than ISIS’s wins (and losses) on the ground in recruiting converts for the cause. The United Kingdom has put a lot of work into removing terrorist content from the Internet. However, the U.K. remains a major recruiting base for ISIS. It’s hard to evaluate the success or failure of programs such as this, and it could be that the U.K. situation would actually be worse in the absence of these censorship efforts. However, there’s no real evidence to support this theory. Other proposed countermeasures — such as responding with counter-narratives and counter-propaganda — don’t have a demonstrably better track record. While Fidler doesn’t say this, it is plausible that policymakers fixate on Internet-based responses to ISIS recruitment strategies because they are visible, politically salient and relatively cheap. Unfortunately, this does not mean that they are effective.

Censorship sits awkwardly with free-speech values

Fidler points out that both public and private efforts to remove terrorist content can clash with free-speech values. This is most obviously a problem for the United States, which has been highly reluctant to engage in censorship, given the willingness of the Supreme Court to strike down laws that could impinge directly on free speech. Accordingly, the United States has effectively delegated out much of the job of censorship to private businesses such as Google’s YouTube and Facebook, which are allowed (and indeed encouraged by laws such as the Communications Decency Act) to have their own codes of conduct over which content they will host and which content they will take down. The government, furthermore, sometimes requests that companies take down content. In principle, these requests are nonbinding, but in practice they may be hard for businesses to ignore.

As Fidler indicates politely, this has created a terrible mess. There is very little transparency around government requests for censorship (by coincidence, the Electronic Frontier Foundation has just published a short piece complaining that Facebook does not provide any information on U.S. government censorship requests). In addition, different private actors, with different codes and standards, engage in private forms of censorship on their own behalf that generate confusion and inconsistency. Companies have to make complicated judgment calls. For example, I’m aware from my own conversations that YouTube initially censored footage from pro-democracy protests in Iran in 2009 showing the violence perpetrated against the protesters, on the grounds that this violated its code of conduct. After thinking through the political implications of this censorship, YouTube changed its mind — but it could (with different leadership) have opted for the opposite choice.

More transparency might help, but it won’t solve the underlying problem

Fidler proposes a set of modest reforms aimed at introducing greater clarity and transparency. He suggests that the U.S. government should issue a presidential directive setting out the circumstances under which the government will request that private companies take down content. He also pushes for private companies to explain their policies and to get independent experts to review them, and for the government’s Privacy and Civil Liberties Oversight Board to review the requests made by the government and report on them to Congress and the public. As Fidler notes, government and business are under pressure to “do something” about online propaganda, to counter ISIS’s success in recruiting foreign fighters. It’s not at all clear that anything that the government and private business can do will be effective. Fidler doesn’t have any silver bullet solutions to this problem — but he does propose ways to minimize the damage that the patchy U.S. anti-terrorist censorship regime could do to civil liberties and transparency.