Dave Winer today notes a vulnerability in the Twitter system that could allow the advent of the spammers' invasion of the system to occur. It is at least one of the ways I've seen for the system to be easily gamed for the slightly grey-to-black hat inclined. Dave's purported security hole lies within the newly minted replies system Twitter recently rolled out, as he illustrates both in practical and theoretical screenshots today:
Another problem, the destination of the url is likely masked through the use of a shortener so the user could be clicking through to some really nasty place, with no way of knowing in advance that's where they're going. (Such messages probably wouldn't alert you in advance that they're about meds or poker or sex.)
Most Twitterers have the habit of following back those that follow them first, especially if they appear to be someone aligned with them in terms of the topics they twitter about. Adding folks in this manner can add literally hundreds of followers a minute, and one could piece together the code for this in a manner of hours (I know, I've tried).
Security through obscurity never works, at least not for long. Granted, Twitter is a system of trust, as most social networks are, but certain safe-guards are going to need to be thought about and adopted soon, as Twitter is soon to hit the mainstream.