It depends on your perspective. As a roboticist, I view A.I. as a discipline within Robotics, which itself is a discipline of Computer Science. The difficulty in definitively labelling these fields is partly due to the massive amounts of overlap. Take robotics for example, which combines the most advanced elements of mathematics, mechanical engineering, electronics, philosophy, neurology and many more. Do bear in mind that artificial intelligence need not be hosted on silicon (i.e. computer chips), it is entirely possible to have biological artificial intelligences, 'running' on living material. Look at the works of Professor Larry Bull from the University of the West of England (my university professor):
Bull, L. and Uroukov, I. (2008) Towards neuronal computing: simple creation of two logic functions in 3D cell cultures using multi-electrode arrays. International Journal of Unconventional Computing, 4 (2). pp. 143-154. ISSN 1548-7199 Available from: http://eprints.uwe.ac.uk/20720
They where able to demonstrate the logic functions AND and OR using living neurons, harvested from chickens. This type of A.I. research is so different from conventional computer science that it cannot be labelled under the same field