New research reveals the extent that ‘bots’ are infiltrating and manipulating political debate on-line – with the majority pushing extremist views.

Analysis around the Twitter debate about the ‘anti-mask movement’ reveals one in fourteen accounts tweeting about the issue came from an automated bot account – with 69 per cent pushing either hard left or extreme right wing political content.

The research follows a recent demonstration in London that saw far right and the hard left protesters join forces to brand Covid-19 a hoax and to highlight the ‘tyranny’ of coronavirus restrictions.

The report, carried out by Kekst CNC, was commissioned by Mainstream UK, the campaign group against political extremism, as part of its work to highlight how extreme views are being spread across social media.

Chairman Ian Austin said it was important for the Government in its upcoming On-Line Harms Bill, to ensure social media platforms were stamping out on fake accounts as well as targeting real account holders fuelling hatred on-line.

Lord Austin said: “Fake accounts promoting fake news about Covid-19 represent a real danger to public health. They are spreading the bogus claims of the cranks and conspiracy theorists who want us to believe Covid-19 is not a real danger to our health and we should ‘take a stand’ by refusing to wear masks.”

Bots are automated accounts that push certain messages or re-tweet posts to increase engagement on real accounts by artificially inflating the number of social media shares it receives.

They are also used to manipulate and influence public debate on particular issues with swarms of bot accounts being used to amplify certain political messages and drown out opposing views.

Researchers reviewed the authors of all tweets about the anti-mask movement posted in August. Of the 36,974 social media posts, reviewed by Kekst CNC, some 2,588, seven per cent, were generated from bot accounts.

While some 40 per cent of these bot accounts have been active for less than six months they had generated high levels of content. A quarter of the bot accounts had tweeted more than 50,000 times, some had tweeted as many as 300,000 times. In 69 per cent of cases the re-tweets would be sharing content that was primarily about political subjects.

In the conversation around the anti-mask movement, 19 per cent of content shared was driven by bots pushing left wing content while 50 per cent was sharing extreme right wing content. Only 17 per cent came from accounts promoting general content about masks and coronavirus.

Previous research has suggested bot activity accounts for between five and 15 per cent of all activity on Twitter but the company disputes these claims and says it has got better at identifying and removing automated accounts.

Writing earlier this year, Twitter’s head of integrity, Yoel Roth, and Nick Pickles, director of global public policy strategy, said: “Going back a few years, automated accounts were a problem for us. We focused on it, made the investments, and have seen significant gains in tackling them across all surfaces of Twitter.”

However the research reveals the extent to which fake accounts are still fuelling Britain’s deeply divided political conversations on-line and raises questions about whether Twitter is doing enough to combat misinformation on its platform.

Lord Austin said: “Twitter conversations about politics can already been deeply polarised so it is really worrying to see the extent to which divisions are being fuelled by fake accounts.

“These accounts are spreading dangerous misinformation about Covid-19 and helping lend credence to the conspiracy cranks who think this killer virus is all a hoax.

“Twitter does encourage users to report suspected bot accounts and seeks to remove them – during the course of this research five per cent of the monitored bot accounts were taken down. But as fast as they are being removed, new accounts spring up and it is clear Twitter is being out-paced by the size, scale and speed of these automated accounts.

“In one month just 1,000 of the bot accounts being monitored generated more than one million tweets.

“More needs to be done if Twitter is not to become simply an on-line haven for the promotion of misinformation and the manipulation of public debate. ”

Bot accounts are programmed by humans to act in a certain way and the high levels of political content being shared suggests a deliberate attempt to influence how the public view certain controversial issues. The research will raise fresh questions about whether Twitter can be viewed as a reliable source of on-line content.

At the end of August more than 10,000 anti-mask demonstrators gathered in London’s Hyde Park to brand coronavirus a hoax despite tens of thousands of deaths and to protest against the “tyranny” of restrictions brought in to protect the public.

The protestors heard speeches from conspiracy theorists David Icke and Piers Corbyn, brother of the former Labour leader. Icke has already been banned from Facebook for publishing “health misinformation that could cause physical harm”.

Twitter is facing growing demands to do more to clean up its platform. In July thousands of account holders staged a 48 hour boycott in protest over Twitter’s failure to close down the account of grime artist Wiley after he posted a string of highly offensive antisemitic posts.

Notes to editors:

Kekst CNC is a leading global strategic communication consultancy. The team of over 250 experienced professionals serve clients from 13 offices in New York, London, Munich, Berlin, Frankfurt, Paris, Brussels, Tokyo, Seoul, Hong Kong, Dubai, Abu Dhabi and Stockholm. As trusted advisors, the firm contributes its expertise on such high-stake matters as: M&A, shareholder activism and governance, crisis communications, restructurings, regulatory investigations, litigation support, investor relations, IPO communications, issues and reputation management, change management and employee engagement, as well as digital and social communications. For more information, please visit: