Old, but I missed it previously. It's horrifying to read that this kind of thing isn't practiced only against state actors. The risk here is that these techniques (and false news, for that matter) if over-used will inevitably make people doubt the veracity of everything they see online, in emails, and in texts. So the risk they run is that these tools go from profoundly effective to not effective at all. Already I find myself believing less and less of what I read online (for example). I would encourage government agents and agencies to exercise significant restraint in the use of these tactics in order to maintain the effectiveness of these tactics against truly bad actors, and to protect the integrity of the web.
This happened in a big way during the election -- the Clinton campaign poured millions into paid online supporters to go on reddit, etc. through an organization called Correct the Record. Once this became known every post in support of Clinton was suspect; people noticed similar comments appearing out of the woodwork, accounts which did nothing but post pro-Clinton comments. The entire politics subreddit turned into a lockstep echo chamber -- no anti-Clinton dissent of any kind allowed for a month or so before the election. Anyone who mentioned correct the record was instantly banned. I'm sure there was some amount of Record Correcting out there but the tragic part was that every single comment in support of Clinton no matter how earnest and honest and real was immediately suspect and accused of being a paid troll. It's impossible to have a real discussion when you are always wondering if the other side of the conversation is not arguing in good faith, and that's the exact climate this sort of thing creates.
It's basically a Sybil attack on public discourse. Sybil attacks, when conducted for any reason, do in fact almost always lead to the destruction of the network being attacked.
> do in fact almost always lead to the destruction of the network being attacked.
Only if the network relies on each identity corresponding to a "real" person. Some networks, such as Bitcoin, couldn't care less. Some networks can be DOSed, but not very well, and with simple mitigations, like Bittorent.
When climate change discussion was ripe, I saw evidence of mass manipulation via fake profiles on Yahoo news. It was so obvious and disgusting that I stopped reading commentary on news sites and eventually news entirely. I also see patterns of suspicious behavior on Reddit that could be attributed to social experiments by programs such as discussed in the article. I have no doubt political agendas are promoted through such programs all the time. Several years ago it was already as simple as having access to the right forums and you could get your hands on software for FB mass account creation which could be used for herding. It's really scary how much influence a single individual can possess online with the right tools and how easy it is to manipulate people. Imagine having access to 10 million pairs of eyeballs to manipulate as you please - such is possible on Facebook.
