5G mast arson, coronavirus conspiracy theories force social media to walk a fine censorship line

SAS Head of Data Science Iain Brown discusses how social networks can tackle harmful conspiracy theories, while also not stifling free speech.
Written by Charlie Osborne, Contributing Writer

The COVID-19 outbreak has forced many of us to stay at home, work remotely, and more now than ever, rely on our broadband and mobile networks to maintain contact with the outside world. 

From a business perspective, 2020 was due to be the golden year for the launch of 5G mobile devices, following trials in cities across the world to test the groundwork for the eventual deployment of the next-generation wireless technology. 

Access to 5G is scattered at best as telecoms providers have been forced to bring deployment plans to a screeching halt due to COVID-19. While some devices that support 5G -- including the Samsung S20 Ultra 5G and LG V60 ThinQ 5G -- are commercially available, vendors are grappling with supply and production issues while the pandemic lasts. 

5G is also facing a problem that its predecessors, 3G and 4G, did not -- rumors linking the wireless technology to public health, giving rise to a range of conspiracy theories and even arsonists targeting the equipment providing the backbone of connectivity. 

Over social networks including Facebook and Twitter, there is a variety of 5G 'awareness' groups and profiles promoting 5G conspiracy theories. These include:

  • 5G lowers our immune systems, making us more susceptible to illnesses including COVID-19
  • 5G is "untested" and increased deployment, frequency "concentration" causing harmful effects to health
  • 5G is radioactive and destroys local trees and wildlife
  • Coronavirus spreads through 5G waves
  • The coronavirus lockdown is designed to allow for the installation of "weapon" 5G masts
  • 5G "emissions" and antennae have caused the pandemic

In some cases, social media users have gone a step further than the discussion of conspiracy theories, urging the destruction of 5G masts for the sake of public health -- and there are individuals that have obliged.

Pictures: London's 5G trial takes to the rooftops

There have been at least 61 suspected arson attacks against telephone masts in the United Kingdom alone in recent weeks, including against those serving local hospitals, and it is believed the uptake in vandalism is due to the deluge of 5G theories being shared across social media.

The Netherlands, Belgium, Italy, Cyprus, and Sweden have also experienced mast arson attacks. 

While these theories have been debunked, with agencies including the World Health Organization (WHO), technology vendors, telecoms providers, and governments attempting to stop the spread of such misinformation, the conspiracy link between 5G and the novel coronavirus persists. 

See also: AT&T Q1 mixed amid COVID-19, plans more 5G investment as network traffic surges

Earlier this month, the UK government branded the theories as "crackpot" and "dangerous nonsense." The arson attacks, too, have spurred on government officials to hold talks with technology giants to explore how to control the spread of this misinformation. 

However, there has to be a balance between the removal of potentially dangerous content encouraging criminal damage, free speech, and what could be considered to be censorship. Facebook, for example, removes pages and groups urging the destruction of 5G equipment, while YouTube attempts to restrict search engine results relating to this form of content, but it is still easy to find 5G/COVID-19 conspiracy fake news and discussions. 

The problem, according to Dr. Iain Brown, Head of Data Analytics at SAS UK&I, is there is no "one size fits all" approach to moderating this kind of content -- even with the benefit of computer science to take some of the pressure away from manual labor forces. 

In an interview with ZDNet, the executive said that social networks differences -- ranging from WhatsApp's encryption to Facebook's massive ad network -- require a variety of moderation approaches, but every network is able, and should, take advantage of automation to at least understand how misinformation appears on the platforms. 

SAS is one of many vendors working to develop network analysis and pattern detection across multiple platforms and industries, with content moderation being one application of Artificial Intelligence (AI) and Machine Learning (ML).

"This means finding pre-classified fake news examples to create good quality training data for AI machines to learn from," Brown said. "This, in turn, will help a machine learn how to spot patterns and flag misleading or harmful content. Social media companies should observe a fundamental principle of data science -- only good quality data can produce good quality insight."

CNET: All about 5G: What it is and why you'll want it

AI and ML have been used to bolster content moderation efforts on social media networks for years, but these systems have been known to demonstrate bias, not only against individuals but also against free speech

There is a concern, then, that 5G conspiracies -- albeit debunked and potentially leading to vandalism -- as a borderline topic, if banned entirely, could be eroding the right to demonstrate an opinion. 

Ethics, too, needs to be considered to stop AI and ML technologies from venturing too far down the censorship rabbit hole. Brown told us that current solutions can make good judgments, but are "not yet advanced enough to make perfect decisions all the time."

"In the interest of ethics, it's vital that humans make final judgments, to avoid jeopardizing free speech," the executive added. "It's good to see that many social media companies are already doing this, employing both humans and machines on the front line of moderation."

In light of COVID-19, however, social networks are also facing a trade-off: under pressure and reduced workforces have led to the question of whether or not to tighten or relax rules to maintain free speech while also removing harmful content. 

TechRepublic: 5G mobile networks: A cheat sheet

This conundrum, the executive says, is a challenge that might require changes to automatic content moderation systems -- at least for now -- but human moderators still need to be at the heart of decision-making.

"In a crisis like this, social media companies must be careful not to inhibit the flow of conversation," Brown said. "It is important that citizens are able to engage with the topic and exercise the right to free speech whether through commentary, satire, political discussion, or any other means. Social media corporations must consider the balance of time and accuracy in their moderation processes, keep on top of what fake news looks like, and act fast to protect against the harm misinformation can cause."

Top programming languages, 5G worries, cloud computing, and more: Research round-up

Previous and related coverage

Have a tip? Get in touch securely via WhatsApp | Signal at +447713 025 499, or over at Keybase: charlie0

Editorial standards