The GDI is concerned about how different disinformation actors are abusing the internet ecosystem and the harms that this creates for our societies, economies and political processes. To break the network and mitigate these harms, we need policies that understand how the model of disinformation operates. We need to strip down the network to its basic parts and track who is spreading disinformation, how, and why.
This has led the GDI to recently undertake research and present a different approach to explain disinformation as a network phenomenon.
Based on these findings which we have just published here, we see a new model for disinformation in which it:
- operates across multiple platforms,
- is driven by a spectrum of threat actors,
- cloaks seeds of truth in conspiracy, and
- uses language and content to form narratives that fan the flames of conflict.
As we argued earlier on this blog, disinformation – or adversarial narrative conflict – is a form of fifth-generation warfare (5GW) that is carried out across a wide variety of platforms, from Facebook to Amazon to Voat.
For the GDI, adversarial narratives are defined as:
“Intentionally distributed narratives without a required chronology or sequence of content (“artefacts”), and which seek to enrage and divide internet users”.
When adversarial narratives are deployed, they create a series of smaller conflicts played out across the internet.
We see this clearly in the adversarial narrative conflict that has emerged on the issue of fifth-generation (5G) mobile networks in the US and elsewhere.
As we profiled here on the GDI blog and in our research, the “stop 5G” narrative is based on some factual legitimacy: it dates to comments about its regulation originally made in 2016 by the then Chair of the US Federal Communications Commission (FCC).
But as the narrative has travelled downstream, fabricated and conspiratorial layers have been added on and linked up with other disinformation efforts – including the Flat Earth Movement.
This has happened even as the narrative has gone more mainstream, like getting amplified by RT and even getting picked up by Fox News, in the case of 5G. The recruitment of these more “mainstream” sources has effectively led to a conspiratorial “red pilling” of the narrative to the broader public.
This model of disinformation is critical for understanding how we measure disinformation risk and work to upend the underlying ecosystem. To be effective, we must have policies that operate across platforms and push for collective, coordinated responses. Otherwise, the current content and platform policies will continue to be ineffective against the problem.
Below are some preliminary recommendations for advancing work to diagnose, identify and prevent the new nature of disinformation campaigns:
- Agree to rapidly develop initial working definitions, examples, and frameworks for harmful online content for use by platforms and the public.
- Promote robust collaboration on content moderation across social media platforms, cloud service providers and ecommerce providers, among others.
- Share threat intelligence (using harmful content frameworks) with key stakeholders from the platforms, civil society and government (i.e. the ISAC/ISAO model).
- Get technology companies and governments to commit to fund, develop, and promote counter-messaging campaigns through the consultation of former members of fringe communities.
- Ensure platforms, brands and ad exchanges work together to demonetise and de-fund disinformation actors and their domains.
These policy recommendations aim to move away from singular actions and misdiagnosed responses towards collective actions that break up the disinformation ecosystem.