To some degree this is presumably a legal question, but my question is really trying to ask about the principle here.
I guess the core question is about what expectations a "reasonable" agent might be construed to have as regards privacy.
The edge case where even consciousness is somehow public and legible is what particularly interests me here, though. What if any might the ethical implications be of inspecting a sentient machines' code, or perhaps more provocatively, somehow directly inspecting its "qualia" while it's running?
(Though presumably at this post-Singularity point, many of the ethical referents that sustain an individual expectation of privacy would have mutated irrecognizably...)