Read on Twitter
This seems a good moment to highlight the importance of keeping perspective and sticking to the evidence.
A few thoughts on "bot networks" in general, and Chinese info ops in particular. https://twitter.com/donie/status/1258861199862779922
A few thoughts on "bot networks" in general, and Chinese info ops in particular. https://twitter.com/donie/status/1258861199862779922
First point: without seeing the evidence, we can’t assess the claim of “Chinese bot networks”.
But the best way to ensure the integrity of the information space is to be super careful to verify the evidence.
Remember, in this space, evidence is everything.
But the best way to ensure the integrity of the information space is to be super careful to verify the evidence.
Remember, in this space, evidence is everything.
That's especially important because disinfo actors deliberately target the very concept of evidence, and try and sow panic and confusion by obscuring the facts.
Judging things according to the evidence is one way not to play into their hands.
Judging things according to the evidence is one way not to play into their hands.
That said, what do we know about bot networks generally?
They’re the pawns on the disinformation chessboard. Small, cheap and disposable.
You used to be able to rent them for a fistful of dollars. (h/t @josephfcox) https://www.thedailybeast.com/i-bought-a-russian-bot-army-for-under-dollar100
They’re the pawns on the disinformation chessboard. Small, cheap and disposable.
You used to be able to rent them for a fistful of dollars. (h/t @josephfcox) https://www.thedailybeast.com/i-bought-a-russian-bot-army-for-under-dollar100
That’s because botnets are fundamentally a pretty basic idea: a bunch of centrally controlled accounts that retweet, like or post prepared texts.
Twitter cracked down on them pretty hard in 2018. (h/t @JonRussell) https://techcrunch.com/2018/02/22/twitter-is-finally-cracking-down-on-bots/
Twitter cracked down on them pretty hard in 2018. (h/t @JonRussell) https://techcrunch.com/2018/02/22/twitter-is-finally-cracking-down-on-bots/
The thing is, fake accounts, especially new-born fake accounts, *don’t have real friends*.
(Sniffs into handkerchief.)
Lacking real followers, even if they retweet stuff, who's actually going to see it? Only the account they retweet, in its notifications.
(Sniffs into handkerchief.)
Lacking real followers, even if they retweet stuff, who's actually going to see it? Only the account they retweet, in its notifications.
Now there can be uses for that. I've seen botnets used for harassment and intimidation before now, tagging someone and mass-retweeting it.
That's a niche use case, though. Not nice, but niche. https://medium.com/dfrlab/botspot-the-intimidators-135244bfe46b
That's a niche use case, though. Not nice, but niche. https://medium.com/dfrlab/botspot-the-intimidators-135244bfe46b
In general, a bot-blast can increase the number of retweets on a post or the followers of an account, and that can make it look more authoritative.
But that’s a cosmetic gain, and if too many bots do it at once, they risk getting caught by the automated systems.
But that’s a cosmetic gain, and if too many bots do it at once, they risk getting caught by the automated systems.
So if China *is* using a bunch of bots to boost followings and amplify (and we haven’t seen the evidence), that’s a high-volume, low-tradecraft effort with minimal likely impact.
It’s worth comparing that with other known pro-China networks that have been analysed recently…
It’s worth comparing that with other known pro-China networks that have been analysed recently…
Last August, Twitter took down a “large, spammy” network targeting Hong Kong.
A lot of the accounts looked like commercial spammers. High volume, low tradecraft, low impact. (h/t @DFRLab) https://medium.com/dfrlab/twitters-hong-kong-archives-chinese-commercial-bots-at-work-f4c7ae8eea64
A lot of the accounts looked like commercial spammers. High volume, low tradecraft, low impact. (h/t @DFRLab) https://medium.com/dfrlab/twitters-hong-kong-archives-chinese-commercial-bots-at-work-f4c7ae8eea64
“Spamouflage Dragon” worked across platforms, using a mix of new accounts and hijacked ones.
It didn’t even clean out the hijacked accounts’ early posts before it shifted to Chinese posting.
High volume, low tradecraft, low impact. h/t @Graphika_NYC https://graphika.com/reports/spamouflage/
It didn’t even clean out the hijacked accounts’ early posts before it shifted to Chinese posting.
High volume, low tradecraft, low impact. h/t @Graphika_NYC https://graphika.com/reports/spamouflage/
Of course we don’t know what undiscovered networks are out there, but the covert Chinese online efforts we’ve analysed so far have been low-quality and low-impact.
That’s not to dismiss China as a threat actor, but it’s a reminder that perspective and evidence are vital.
That’s not to dismiss China as a threat actor, but it’s a reminder that perspective and evidence are vital.
Remember, info ops often try to sow panic. They want to look more scary and effective than they really are, because people who are scared or angry are easier to manipulate.
Stick with the evidence. Keep it in perspective.
Let’s all stay calm.
Stick with the evidence. Keep it in perspective.
Let’s all stay calm.
