TMS, when using approved devices on participants who comply with the mandatory inclusion/exclusion criteria, has an excellent safety record. However, TMS does not "deactivate" the brain as such. The addition of the stimulus either acts as noise to disrupt the pattern of activity; artificially induces an action potential; or when used in a repetitive train sense alters the resting potential of the neurons in question. Also useful to remember is that TMS does not just affect the area immediately underneath the coil, it will have an effect on the network of functionally connected neurons as well.
The greatest chance of damage would come from poor experimental design and management. For example, a patient with epilepsy (or a participant with a family history of epilepsy but no current or previous diagnosis) could suffer a seizure, which might lead to brain damage. It is always worth noting what medications a participant is taking, whether or not they use recreational drugs, and whether or not they had been drinking (moderately to heavily) in the 24 hours prior to the session.
The short answer is that TMS is safe when used sensibly in a well designed and controlled experiment/treatment.
Please take time to familiarise yourself with the mechanics, physics, physiology, and considerations surrounding this method. The oxford handbook of TMS, or the TMS a chronometrics of mind, are both brilliant starting places. Likewise, Magstim offer regular training courses, and other educational events that are invaluable to attend if you are new to this technique.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3260536/ A "recent" report on ethics and safety associated with TMS.
Indeed, I agree with the above answers based on my 10+ years of TMS experience. Note that TMS, using very specific paradigms, has also been used to prevent/reduce epileptic activity, although a history of seizures should still be considered a contraindication to TMS.
TMS can be used to deactivate brain activity i.e. virtual lesions, but the design of the study needs to be clearly planned and meticulously executed with careful consideration of cortical overflow.
Yes. Using particular frequencies (of light) can achieve this. Type in "optogenetics inhibition" in Pubmed and you will find many publications related to this.