Given the proclaimed state of research (e.g. IBM's Blue Brain) one could deduce a necessity to regulate the affairs surrounding the creation of artificial intelligences. Deciding on very restrictive laws in retrospect could be seen as inappropriate and may be much more difficult than early regulation. Regulation could turn out to be very time-consuming, given the world-wide need for regulation and the complexity of the topic.
Extending human legislation to artificial intelligences could also be seen as common courtesy towards any non-human intelligences one might happen to come across. ( If one assumes that life doesn't develop naturally in hyperspace any denizens of hyperspace are quite likely to be AIs. )
A plausible inversion of trying to sue an AI in absentia would be to grant an AI protection under the same basic rights as they apply to humans. Both would be "extending human legislation to AIs". As a precedent it may be one precedent required for AIs to do the same.