What if all your friends were really chatbots and spies?

If you bother following tech news at all, it’s really no wonder that tinfoil hats are so fashionable these days. We recently talked about the takedown of security firm HBGary by the internet activist group Anonymous, but I didn’t bother mentioning one amusing thought that kept occurring to me, and frankly should occur to anyone who applies any thought to whistleblowing, reputation management, and disinformation strategies. Which is: what if HBGary was just a sort of honey trap for hacktivists? What better way to infiltrate your opponent than allowing them to think they’d infiltrated you, and made off with 4GB of sensitive internal e-mails? Four gigabytes of e-mails loaded with discussions about security exploits, with file attachments that were themselves exploits? What a great Trojan Horse to deliver to your enemy! And while perhaps I’m being a bit flip suggesting such a thing, you should find the reality of what has been revealed by perusing the HBGary e-mails even more disturbing. I personally have a lot of friends that I rarely see but often banter with online, and often joke that I’m just a sophisticated chatbot. Which for now of course is an absurd notion; in spite of supercomputers beating humans at Jeopardy, the average chatbot is still in the “conversation with a dullard” phase of development. But ponder this: so are most humans who spend their time commenting all over the internet. So here’s the disturbing bit. By now you should be familiar with things like political astroturfing and internet sockpuppets, but this DailyKos piece discusses what may be an even more disturbing concept that HBGary was working on – “persona management” software that allows the user of the software to appear online as an army of commenters to manipulate opinion and erode online trust, much like Digg Patriots.  It’s questionable how effective the tool would be at this point in terms of creating credible personas, but what is actually more worrying about this kind of tool is the latter notion. We’re already bombarded daily by various forms of phishing attempts, but what if all our social network interactions, blog commenting, and Twitter/RSS feeds were partially tainted by an intelligently-crafted stream of consensus manipulation, as that DailyKos piece suggests? Oh, and by the way. The US government is an interested customer. I look forward to your comments, even if you ARE just a piece of sentiment-manipulating software.