Reuters discovered that the innovative voice recognition feature converts swear words into a string of # symbols when rendered into text.

Why the censorship? A spokesperson from Google says the “feature” is less about sanitizing users’ speech and more about making sure curse words don’t accidentally appear in transcriptions erroneously, a real possibility given the early and not always accurate state of voice recognition technology.

“We filter potentially offensive or inappropriate results because we want to avoid situations whereby we might misrecognize a spoken query and return profanity when, in fact, the user said something completely innocent,” the statement said.

Fair enough. Although that seems to imply that one day, when the technology improves, Nexus One users may see their right to be foul-mouthed returned. One can only ###ing hope.