Social media ‘bots’ used to boost political messages during Brexit referendum

In summary

First-of-its-kind research by Dr Marco Bastos found how a network of social media ‘bots’ – automated accounts – were used on Twitter to artificially amplify electoral messages during the 2016 Brexit referendum campaign.

The researchers found that 13,493 accounts tweeted on the two weeks before and after the referendum before disappearing after the polling stations closed.

The work received widespread media coverage and has contributed to UK Parliamentary inquiries and subsequent legislation on the need for social media companies to police disinformation. It has also been used by security services to develop responses to misinformation campaigns worldwide.

What did we explore and how?

City research want to make transparent potential harms caused by social media and the use of algorithms or non-human accounts. These methods have been used to spread misinformation and interfere with democratic processes.

Using user activity metrics and temporal posting patterns, research of Twitter during the 2016 UK referendum on EU membership found a ‘Twitterbot’ network that was active during the campaign but closed after its completion.

These false online identities voiced opinion and attempted to manipulate public opinion through rapid cascade tweets.

The author identified how the ‘botnet’ was subdivided into specialised subnetworks dedicated to ‘echoing’ through retweets created either by other bots (driven by algorithms) or humans.

The authors could show how ideologically polarised ‘echo-chambers’ on Twitter mapped onto geographically situated social networks. This allowed for the creation of an algorithm that could identify user location and political affiliation.

Follow-up studies found that a third of the messages leading up to the referendum vote were removed and that only about half of the most active accounts during the referendum continue to operate publicly, with a fifth of these accounts taken down by Twitter.

The universe of removed messages also skews towards leave. Indeed, there are more messages from the Leave campaign that have disappeared than the entire universe of tweets affiliated with the Remain campaign.

Benefits and influence of this research

The discovery of the Twitterbot network and the consequent public policy implications impacted in three ways:

1. Contributing to UK Parliamentary inquiries into the role of social media networks in dis/misinformation and the resulting changes in public policy leading to draft legislation and developing regulatory frameworks.
   • This work resulted in oral evidence given to a House of Commons Digital, Culture, Media and Sport Select Committee, influenced the Government’s Online Harms White Paper and set out a ‘duty of care’ for social providers.
2. Enabling national and international security organisations to understand botnets and the threats posed by social media to the democratic integrity.
   • The research informed Parliament’s Intelligence and Security Committee, was mentioned in parliamentary debates, and referenced in a House of Commons Briefing Paper on ‘National Security and Russia’.

     Dr Bastos spoke at a joint Royal United Services Institute for Defence Studies / UK Cabinet Office expert round-table discussion, and informed meetings with NATO’s Strategic Communications Centre of Excellence and the UK Foreign & Commonwealth Office.

3. Bringing these issues into the public spotlight and making visible how social media platforms can be abused and social media companies operate.
   • Between October and December 2017, 220 national and international media outlets reported on the research. As a result, Twitter invited City to apply to its research programme for funding from a new scheme they had instigated into election integrity.

     Dr Bastos won £90,000 for research on ‘the Brexit Value Space and the Geography of Online Echo Chambers’.