Current developments in genomic technology challenge the traditional normative framework for biomedical research and its well-known components. It has become clear that the common interpretation of the concepts of privacy and confidentiality as being absolute or near absolute cannot be sustained. Whenever genetic samples are involved re-identification will be possible. Although the research community is well aware of the facts, until now this awareness has not been reflected in the language of consent. Therefore, in many cases, existing consent cannot be assumed to be fully valid.
Although I love him dearly, I think Steve has overstated what George Church is proposing with the PGP. No one I know of is advocating immediate wholesale abandonment of the current model; the idea is simply to begin to integrate other informed consent approaches into human genomic research. Of course, whether the powers that be buy into that notion is another matter.