The global spread of misinformation has been credited with supporting a string of shock political results, from Donald Trump’s US election victory to the UK’s vote to leave the European Union.

And while it’s impossible to accurately calculate the extent to which fake news shapes public perception, there’s no doubting it does have influence. Analysis by the Oxford Internet Institute has shown that during the run-up to the US election, professional news and junk news was shared on social media at a ratio of 1:1. This compares with a ratio of 3:1 (professional to junk news) in the build-up to the French election.

Early estimates suggest the balance of professional and junk news social media sharing ahead of the UK election was likely to have been somewhere between these two ratios. Misinformation is harmful to the democratic process, but what can be done to stop this fake news phenomenon?

In many ways, artificial intelligence (AI) helped cause the outbreak of fake news – but a fightback is underway and AI is a big part of the cure.

What is fake news?

Stories used to be classified as either news or ‘not news’ (ie, puff pieces or light-hearted articles that aren’t deemed informative enough to be ‘proper news’). But in the age of social media the distinction has become far less clear.

It can often be difficult to decipher if news shared via a Facebook post or Tweet is genuine or fake. Worryingly, junk news is also beginning to filter through to the mainstream media.

Speaking at a Tata Consultancy Services technology event, Philip Howard, Professor of Internet Studies at the Oxford Internet Institute (OII), explained what is meant by the term ‘fake news’.

“Fake news is often a story with a few nuggets of truth that becomes what we’d all recognise as a commentary essay,” he said. “The commentary is the stuff that we call fake news.”

Professor Phil Howard talking at TCS Sparks Salon at Hay Festival 2017 Hay-on-Wye Powys Wales UK

Fake news is written and published with the intent to mislead its audience – this is usually motivated by political or financial gain.

How does fake news spread?

There are many ways in which fake news is presented to the masses. One of the most common methods of spreading what the OII refers to as ‘computational propaganda’ is through social media ‘bots’.

Bots are essentially small chunks of code that automate the dissemination of information over social networks.

Source: Tech Crunch

Some bots are harder to spot than others and in certain cases they are used by large organisations, or even governments, to wilfully spread misinformation.

When civil war was breaking out, the Syrian government designed a bot to hijack a hashtag that was being used by journalists to get a grasp of what was happening in the country.

Professor Howard explains: “The Syrian government hired a firm in bot making to flood the Syria hashtag with soccer scores, soap opera stories and tourism pictures to effectively choke off the conversation about what was really going on.”

Bots can also be used to plug a particular political cause or entice consumers to buy certain products.

Stopping the spread of bots will go a long way towards winning the war against fake news.


Source: International Federation of Library Associations and Institutions

An AI for the truth

Technology gave birth to the current rise of fake news, yet it could also help control it.

While education is of course important in ensuring people aren’t fooled by fake news, the best way of ending its influence is to stop it at source.

Cognition co-founder Tabitha Goldstaub, speaking alongside Professor Howard at the TCS event, says AI “could be the solution to this fake news problem”.

To back this up, she points to efforts that have been made by the likes of Google and Facebook to clamp down on junk news.

Tabitha Goldstaub speaking at TCS Sparks Salon at Hay Festival 2017 Hay-on-Wye Powys Wales UK

The latter has designed a machine learning clickbait detector that identifies and demotes possible fake news stories, while the former has unveiled changes to combat the appearance of fake news in search results.

Meanwhile, Google has also dished out $168,000 to three UK start-ups that are working on AI-based fact-checking projects to assist journalists.

One of the recipients of the Google grant, Fact-checking charity Full Fact, has put its $56,000 towards what it claims will be “the first fully automated fact-checking tool”, which will alert journalists to inaccurate claims made online.

A future we can believe in?

Using AI to stem the flow of fake news is going to be expensive so it’s unlikely that widescale deployment of junk news monitoring will be rolled out until there’s a real business need for it.

But after recent election results, the pressure to tackle the problem has risen significantly. AI will be a powerful weapon in the fight.