I got p0wned earlier today by the same sort of chat-bot/spam exploit I've been seeing from some of my friends.<p>As a Chrome user on Linux, and a pretty much lifelong user of Linux on the desktop, I am rather unaccustomed to being the victim of such exploits, so I didn't immediately know what to do. This one appears to be purely browser/JS-based and/or perhaps exploits some weakness in the Facebook API.<p>It started when a (presumably "infected") friend of mine posted on my wall. It looked to be just text, but presumably contained a trigger for this exploit. Anyway, within seconds, somehow, unbeknownst to me, I was apparently initiating chat conversations with every friend who was online "asking," "Do you have a second?" When they would reply "yes?", I would blast them with some bullshit quiz/test site link, which I can only assume is a phishing farm.<p>Anyway, this continued relentlessly so long as I was logged into the site (and possibly when I wasn't, never definitively established that) until it occurred to me to change my Facebook account password, after which it - knock on wood - seems to have stopped.<p>Does anyone have any idea how this exploit works? It caught me rather off-guard because I expected that sort of thing to be the work of viruses and/or malware on Windows. I would guess that my password was somehow phished out, after which some foreign agent logged into the Facebook messenger as me externally (quite possible to do, numerous IM clients now support the Facebook messenger protocol) and went nuts, but I can't be sure.