Informed consent theory has traditionally addressed the rights of patients and experimental research subjects with the focus in this medical context being retaining the person's autonomy or right to autonomous action under specific circumstances where without informed consent grave wrongs to that person could be done. In this thesis, I argue that this is not an appropriate approach for information technology, where there is often a large number of consent requests; consent requests that are highly manipulative, misleading, or disguised; and a low level of user understanding about computer function and rights and responsibilities of the user and the companies involved.
With our reliance on computers to deal with our personal information and carry out tasks for us increasing at a rapid rate, the ways that we deal with user and software manufacturer rights and responsibilities with regard to software use need much closer examination. There are very few informed consent procedures currently in information technology, and those that exist are grossly weighted toward simply absolving companies of legal responsibility rather than being concerned with the values of the computer user. An appropriate, solid, easily applicable theory for informed consent is required to improve the current ``free-for-all'' situation in computing.
In this dissertation, I examine the history and theory of informed consent procedures in the medical and clinical research fields, and how these have been applied to information technology. I discuss in particular the ``effective consent'' model of informed consent as described by Faden and Beauchamp (1986) that has been adopted as the de facto accepted method for acquiring informed consent from computer users. I then discuss why this approach is inappropriate, looking at several case studies from the areas of privacy and use of software to illustrate the problems with this current theory's application, since, as currently practised in information technology, informed consent procedures place the burden of understanding, assessment of competence, and evaluation of voluntariness, as required by Faden and Beauchamp, firmly on the consenter. The consent-requester is, currently, only responsible for disclosure. The disclosure itself is problematic as well, as often too much information is disclosed, and important information hidden amongst irrelevant or unimportant information or deliberately obscured. Often the information is not applicable to the consenter, or so vague as to be difficult to follow, or often, as in the case of software use and End User License Agreements, written in a language loaded with legal terms and constructions that mean very little to the consenter, or is aimed at further obscuring of important information.
Theory of informed consent is usually focussed on the autonomous decision of the person involved, whether patient, research subject, contract signer, or any other role that requires consent from that party. The idea of autonomy being central to, as well as justifying, the need for informed consent, is almost doctrine, with most major texts on informed consent theory concentrating on the value (Faden and Beauchamp, 1986; Appelbaum et al., 1987). In this thesis, however, I argue that autonomy is not a useful underlying principle on which to justify informed consent in information technology, similarly to Manson and O'Neill (2007), who argued against using autonomy as a justification for informed consent in bioethics. I instead adopt their idea of informed consent as a waiver of normative expectations, exploring Manson and O'Neill's arguments for this approach and their applicability to information technology situations. This is further developed into a solid theory of informed consent for information technology, focussing in particular on the issues in information technology that make an autonomy-centric approach inappropriate and practically infeasible.
Finally, I propose some standards and mechanisms by which this theory can be practically implemented. Mediated by a trusted third party non-profit organisation, I recommend a set of standard license agreements with modular clauses that relate directly to particular normative expectations to be waived by the consenter. These would be incorporated into a new display mechanism that allows users to easily identify the modules through a set of labels that show the concepts encapsulated by each included module, allowing for easier recognition and understanding, and a more informed consent decision.