Election Disinformation in the US
ISD and CASM have leveraged Beam over the course of three years to identify and inform the response to disinformation and weaponised hate online targeting two US election cycles: the 2020 presidential election and the 2022 midterms. This research has raised awareness of online harms among decision-makers in government, law enforcement, community groups, technology platforms, and the public. The project also aims to build an evidence base of gaps in platform policy and the enforcement of those across different types of harmful activity.
To this end, Beam has been deployed as a data-driven detection capability for disinformation, platform subversion and coordinated inauthentic behaviour (CIB) across nine social media platforms, combined with qualitative monitoring of threat actor channels. In order to deliver the insights in an actionable way and in real-time, this research is conducted in partnership with over sixty stakeholder organisations, who corroborate, verify and respond to information threats once detected. This stream of work often acts as the early warning system for information campaigns, online threats and new narratives, targets or tactics in the disinformation and extremism domains. At crisis points, Beam also becomes a vital source of intelligence on online mobilisation for real-world action, including mass demonstrations and targeted violent attacks.
Over the course of this project, we have produced over 45 public research reports that expose the actors and platforms involved in promoting disinformation and hate online. These have covered topics such as election denial conspiracy theories, public health and vaccine misinformation, and analysis of foreign state actor activity targeting American social media users. This program has continued throughout 2022 and in the lead-up to the midterm elections using the same coalition-based model.
ISD and CASM have continuously iterated the methodologies and capabilities available for this work, which leans heavily on the flexibility of the Beam environment to detect and analyse incidents of online manipulation. This particular use case has created a collaborative practice by opening up complex data science techniques to mixed-skills teams of investigators, researchers, and subject matter experts.
