Unpredictable risks and the autonomy of the Internet user

AUTHOR

Lotte Asveld

ABSTRACT

The bankruptcy of the Dutch glass fibre network company KPNQwest in the spring of 2002 raised questions about the acceptability of risks related to digital infrastructures. KPNQwest appeared to be of vital importance to the European Internet traffic since 40 % of the traffic went via the networks of KPNQwest. The consequences of the bankruptcy of KPNQwest were unpredictable, either because of a lack of technical knowledge or because of the sheer incomprehensibleness of the entangled interactions of very diverse actors. According to some, the fall-out of KPNQwest-network would cause a total blackout of the European Internet traffic, according to others no effect would occur at all. Medium estimates amounted to congestions of Internet traffic in different orders of severity.

Since Internet is becoming of increasing importance to our society and more and more people are depending on digital networks, the badly understood risks accompanying the organisation of digital infrastructures are of growing moral concern. The main moral concerns here are related to the notions of autonomy and justice. Autonomy with regard to risk requires that the subject is able to make well-informed choices about the acceptability of a risk. Justice requires that people are treated fairly as equals and no unjustifiable risks are imposed on them. To justify a risk, it is necessary someone take responsibility for the imposition of a risk, preferably the initiator of the action at stake. It is unjust to let innocent subjects bear the consequences of risks they themselves were never able to influence. Moreover, the imposition of the risks can be deemed acceptable trough (hypothetical) consent of the persons affected by it.

Clear possible chains of events or clear risks come with clear responsibilities. Unclear possible chains of event make it easier for responsibilities to be shoved around by divergent parties. Internet Service Providers (ISP’s) rely on the services of KPNQwest and end-users rely on the services of the ISP’s but just how much each relies on the other for the continuation of Internet traffic is unclear. The flow of Internet traffic is hard to register and so are the dependencies between divergent parties. Now whom -if anyone- should make sure there are enough alternatives available in case of a bankruptcy so the flow of data through Internet is ensured?

In the case of KPNQwest eventually other network providers stepped in quickly to fill the hole. Information flows were quickly diverted to other glass fibre-routes. In the end, the bankruptcy of KPNQwest sorted very little effect on the digital infrastructure. The question remains however: was this just a lucky occurrence of events or is the risk of Internet failure through bankruptcy of a provider not so big after all? The fact that nothing serious occurred does not prove nothing could have occurred. In fact, it seems quite probable that any definite answer to the previous questions will never be found until something goes terribly wrong.

The insecurities related to both risks and responsibilities make it hard to guarantee autonomy and justice for the average Internet user. To what extent can Internet users manage the risks they face if they have no insight into what these risks constitute and who is responsible for what?

However, the insecurity about the effects of a bankruptcy of one of the largest infrastructure providers in Europe can also be valued in a positive way. Firstly, because of the insecurities, the trustees in the bankruptcy employed a very cautious strategy. Normally they would have ‘pulled the plug’ as soon as possible because it is a costly business to keep the network running. Now instead, out of fear for claims, the trustees decided to keep the network running, thus allowing more time for customers of KPNQwest to find alternatives. Secondly, many actors within the Internet economical arena: ISP’s, network providers as well as end-users, rather rely on trust than to be forced to put a lot of time and effort into transparency and well-informed choices. Intransparency is time-saving and it does also hide one’s own weaknesses from the competitors.

The main questions in this paper therefor are: should the insecurities related to the reliability of Internet traffic be reduced considering values of autonomy and justice and if so, in what way? Possibly, the clash between autonomy, justice and insecurity is not so hard to reconcile.

One option to deal with these matters is to increase the control governments have on digital infrastructures. Governments could perhaps guarantee the availability of digital infrastructures. Top-down control may however impair the value of autonomy even more than the anarchy that seems to rule Internet traffic nowadays.

To rely on economic incentives is another option. But can the maintenance as well as the development of digital infrastructures be left to the forces of the market? Is it just to trust on the workings of the ‘invisible hand’ as proposed by Adam Smith if many divergent interests depend on the specific facilities?

The notion of Informed Consent may provide a valuable asset here. This instrument can provide a worthwhile interaction between providers and costumers in dealing with unclear risks and expectations. Through this process responsibilities may be openly negotiated and agreed upon. This option is justified by both the principle of justice as that of autonomy.

To ensure a realistic solution to the above-mentioned problem several viewpoints from stakeholders in the process surrounding the collapse of KPNQwest will be taken into account. Stakeholders include clients, but also facility providers (electricity) for KPNQwest and representatives for end-users. Their insights into responsibilities, the reliability of Internet and the need for transparency will provide arguments for determining the ethically most desirable approach to the problems mentioned above, aside from ethical considerations.