I wrote for @GNET_research and @CSCRCoz about how we really don't understand digital-first conspiracy extremism yet. It's different from pre-digital conspiracies, and also different from digitISED extremism which begin offline and move online as a recruiting and comms strategy.
Much more research, and also simply more time, is needed to understand the dynamics of extremist conspiracy movements. In the meantime, however, we can prioritise what we do know:
Based on evidence so far, conspiracy extremists are likely to be lone actors, who have either self-radicalised or been a part of a social media community which has radicalised over time (although this could change in the future).
Not all conspiracies, and not all conspiracy sub-communities, are the same risk for violence. Inherently political conspiracies like QAnon hold a higher risk than crop circle conspiracies, for example. This recognition should help prioritise resources.
The extremist element of conspiracies is only one part of a much bigger picture of harms to mental health, social cohesion and public safety. Like with other forms of extremism, a holistic approach which addresses social causes is needed.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
It's important to caveat that this is not meant to be comprehensive; includes primarily English-language activity; that we have not attempted to verify content; and that we are not seeking to undermine legitimate social media debate. Nonetheless, we made some interesting findings
This included significant spikes in account creation on key days in the conflict, for example Sept 27 and, as well as spikes dating back to earlier clashes in July.
.@AlbertYZhang and I spent the last couple of weeks looking into some coordinated inauthentic activity linked to Chinese-speaking actors. Although small and unsophisticated, it has persistently targeted US and Western audiences on US political and racial issues.
Thread.
The coordinated activity relies heavily on automation (aka bots in the true sense) across both FB and Twitter. We estimate that a couple hundred Twitter accounts (many of which have been taken down) and unknown but likely small number of FB bots have been involved.
The small size of the activity is itself interesting, because it demonstrates how even small-scale actors can engage in attempted foreign political influence. Even if each little campaign has little impact, many small inauthentic efforts like this can blur the information space.
My @ASPI_ICPC colleagues and I analysed a Chinese state-linked influence operation involving 23,000+ accounts on Twitter and related activity on Facebook. Instead of a thread on the report (which you should read!).... aspi.org.au/report/retweet…
I'm going to do a thread on the continuing activity targeting the #GeorgeFloyd#US protests.Some of this is in the report, most of it is newer.
Two caveats up front: first, organic engagement with this campaign was super low. It’s interesting for what it tells us about the priorities and capabilities (or lack thereof) of those behind it, not because it was effective.
When I decided to go down the rabbit hole of digging into one of the websites spreading the supposedly hacked WHO/Gates Foundation/Wuhan lab data, I'm not sure where I thought it would go, but Danish Satanist biohackers was not it.
You guys. This was A TRIP.
Thread.
First, it's worth being clear that it seems very likely there was no recent hack. The data dump is made up of creds scraped from past leaks, and carelessly done at that. vice.com/en_au/article/…
That's why, for example, the WHO file contains creds which have nothing to do with the World Health Organisation, other than containing "who" and "int". Someone just did a "find all" or similar and grabbed everything w/out even checking.
People following Xinjiang related issues on Twitter recently may have seen a few tweets like this. On a hunch that Emma Stone might not be re-tweeting the Global Times’ denial of the repression of Uyghurs in Xinjiang, @tomatospy and I took a closer look.
We found that the accounts obsessively sharing the Global Times article were part of what appears to be a massive spambot network in the making. We identified 375+ related accounts, many of which use the photos of random celebrities.
All of the accounts were created in 2019, but it's not clear who by or for what purpose. Only a relatively small proportion have tweeted on issues relating to China or Xinjiang. Many have barely been used other than what looks like a test tweet - 'Hello [own name]'
My @ASPI_ICPC colleagues @tomatospy@JakeWallis_ASPI and I have spent the last week digging into the data released by Twitter in connection to a state-backed information operation targeting the Hong Kong protests. Here's what we found: THREAD
As initially suspected, the campaign targeting #HongKongProtesters appears to have been hastily assembled with a mix of a small number of purpose-made accounts and a much larger network of repurposed marketing and spam accounts.
In this account, for example, we can clearly see a change of behaviour from 2018 its initial purpose, which seems to have been mostly spamming in English, to a much lower volume of tweets directed at the #HKprotests in 2019.