The smith was working hard on making a new tool. A passer-by looked at his work and remarked that it looked sharp and dangerous. The smith nodded: it needed to be very sharp to do its work. The visitor wondered why there was no cross-guard to prevent the user’s hand to slide onto the blade, and why the design made it easy to accidentally grip the blade instead of the grip. The smith explained that the tool was intended for people who said they knew how to use it well. “But what if they were overconfident, sold it to somebody else, or had a bad day? Surely some safety measures would be useful?” “No”, said the smith, “my customers did not ask for them. I could make them with a slight effort, but why bother?”
Would we say the smith was doing his job in an ethical manner?
Here are two other pieces of news: Oxford City Council has decided to make it mandatory for taxicabs in Oxford to have CCTV cameras and microphones recording conversations of the passengers. As expected, many people are outraged. The stated reason is to improve public safety, although the data supporting this decision doesn’t seem to be available. The surveillance footage will supposedly not be made available other than as evidence for crimes, and not stored for more than 28 days. Meanwhile in the US, there are hearings about the Stop Online Piracy Act (SOPA) and the PROTECT IP Act, laws intended to make it easier to block copyright infringement and counterfeiting. Besides concerns that critics and industries most affected by the laws are not getting access to the hearings, a serious set of concerns is that they would make it easy to censor websites and block business on fairly loose grounds, with few safeguards against false accusations (something that occurs regularly), little oversight, few remedies for the website, plus the fact that a domestic US law would apply internationally due to the peculiarities of the Internet and US legal definitions.
While the purpose of the US bills is certainly not political censorship it is easy to see that the legal and technical infrastructure would make it easy to implement. And just like for current Internet blacklists, it is not hard to imagine mission creep turning the anti-piracy tool into an anti-terrorism, anti-hate speech, anti-climate denial, anti-subversion or anti-criticism tool. It is telling how often critics of internet blacklists tend to end up on them, and the same blocking technology is regularly used today by various regimes to suppress dissenting voices. Similarly it is not hard to imagine that the Oxford cab surveillance data might be made available for more and more public bodies – the vast expansion of public body surveillance in the various updates of the RIPA act speaks loudly that everybody from the police to local councils and ambulance services want access to surveillance data. The initial purpose and limitations are not guaranteed to remain stable, regardless of reassurances.
My smith parable should be pretty obvious: western lawmakers are forging some potentially dangerous tools, and there does not seem to be much concern about making them safe.
Note that I am not arguing that these tools are useless, that is a separate issue. There might well be practical and moral benefits to being able to block certain information or listen in on private conversations. But these benefits better outweigh the risks. In particular, extraordinary extensions of state power require extraordinary safeguards. To not build in strong accountability and serious penalties for misuse from the start is to invite dangerous misuse… and the acceptance of misuse. Oversight organisations need more teeth – and accountability that they do a good job.
From a more abstract level the issue is also how these proposals affect the information processing of society.
Looking at the function of the US acts (and similar proposals elsewhere) their main function is to block access to and trade with certain groups, as determined by certain bodies. The Oxford cab case is another example of increased surveillance powers. Thanks to modern surveillance and information technology more private information can be gathered than ever before. In this case it is about adding another channel of information flow rather than blocking one.
Having the ability to control information flows in society is, as I wrote above, not necessarily bad. There are likely some forms of information that are immoral to spread (like child pornography). There are information hazards where even true information has bad consequences and might need control. Similarly the ability to observe each other can be good for promoting prosocial behaviour and of course sometimes for pre-empting or punishing dangerous or criminal events. The question is how controlling information flows affects society.
David Brin argued that the choice is not between surveillance and no surveillance, but between two-way surveillance societies where citizens can watch concentrations of power, holding them accountable, and one-way surveillance societies where the powerful can observe the citizens but are themselves unobservable. One might make a similar distinction about information control powers: who has power over limiting access to information? In a one-way system a set of powerful groups control who can access what, while in a two-way system citizens have a say too.
At present the power differential in information control is pretty clear. A major corporation or organisation has an advantage in shutting down a smaller group or individual, even when their accusation is groundless. It is unlikely that an individual would get a Telecom or credit card company to pre-emptively shut down a Disney affiliate for copyright infringement if SOPA passes, even if there was good evidence. This is problematic since it puts an extra burden on weak parties when seeking justice or protection, and makes the chilling effects of potential shut-downs unequally distributed. Legislation based on equality before the law and taking vulnerability into account should hence aim to correct this discrepancy.
Would it be possible to have an Internet or even a society if people were free to shut each other out based on perceived or real infringements? It is one thing to have tools filtering what one sees for oneself (freely chosen) to use tools to filter out what others can see. Free speech can function even if few choose to listen but not if voices can be silenced. Hence there are good reasons to be more worried about the US acts than the Oxford cabs: both are dangerous tools requiring accountability, but misuse of surveillance merely entrenches power differences in society while misuse of censorship not only entrenches power, but impairs the ability for an open society to correct itself.
An information society is built on information flows. Control over them gives control over the society. If we want to maintain an open society that means information control needs to be as carefully managed as political and legal control – perhaps more, since we would be impacted by it at all time, and it would affect policymaking directly and indirectly.
The smith making unsafe knives did not have much professional ethics. Lawmakers making unsafe laws might not be much better. And we, as the customers of these laws, should really demand better safety.