top of page

The Online News Association faces disinformation Nemesis

The Online News Association 2019 conference, or #ONA19, held its first session in New Orleans on Thursday, 12 September, and kicked off with a panel highlighting the ambivalent position that publishers face.

On the one hand there is the rapid growth of new media outlets and user-based content, the concept of the individual as information gatherer, photographer, researcher.

On the other - disinformation, lies, propaganda, deceit and hate.

Despite the fact that the main event sponsors are corporate giants like Facebook with its Journalism Project and Microsoft News, the role that these companies are playing in the world of information gathering and promotion continues to pose a major challenge for media owners.

Publishing has been pulverised by these digital juggernauts who are in the business of providing profits for shareholders rather than protecting the freedom of speech. The tension at a previous Online News Association gathering in 2016 which was sponsored directly by Facebook and the journalists on the floor has been replaced by a numbness in 2019.

A kind of post traumatic stress disorder has taken hold of the digital marketplace and the main cause is disinformation.

The Online News Association prides itself in the tenets of truth and justice, yet the reality is digital abuses have become a public health threat as the amount of disinformation spreads and undermines public institutions.

“It’s a losing battle for us because the people who want to share disinformation seem to be winning,” said outgoing ONA board president, Mandy Jenkins.

Two thousand eight hundred delegates from 60 countries around the world gathered at the Sheraton Hotel in New Orleans for the three day conference set that ended on Saturday, 15 September. The tone was set right from the start as a four person panel grappled with this main issue.

“We need strategies and methodologies to deal with these threats,” said Jenkins.

Reuters Global Head of User Generated Content, Hazel Baker explained how deep fake videos were challenging media. “We have a responsibility to get things right,” she said. “Our clients have their own brand integrity to consider.”

Mobile networks are now prevalent throughout the globe and most citizens have access to a mobile phone with a built-in camera.

“These cameras are not being used by professionals. They’re smartphones, carried by witnesses, but how do we know what is authentic?” she asked.

“It’s loaded with risk, but we know that material from users is critical for storytelling and we can’t disregard what is important material,” Baker explained.

At the same time, false information is becoming more complex and difficult to analyse, especially when one takes images and video content into account.

“We have developed a team that has a collaborative workflow, which means having a strategy to deal with visual material,” she said. “You have to know your enemy, so every day we collect and categorise information. The false images are being produced using off-the-shelf applications.”

This means it is cheap and easy for people to create disinformation, but difficult for non-professionals to understand when that material is incorrect. This is the basic strategy being used by organised groups around the world today - making use of international gullibility to spread fear.

Reuters conducted a staged video to test whether people could tell when a Deep Fake video was produced. Most could not tell when a piece of video was computer generated in a process known as Computer Generated Imagery (CGI).

“CGI video takes some degree of expertise to trick people. So how do we train our staff to spot Deep Fake video, when the only examples around when we began this project were academic or from Hollywood and not suitable for frame-by-frame analysis?”

This is why, Baker explains, Reuters created its own video to begin to expose staff to the possibilities using its own actors. They found that two thirds of those who viewed the video had no idea that it had been computer generated. The decision was taken to deal with a neutral topic so as to avoid the possibility that nefarious individuals wouldn’t seize the video and use it themselves to spread lies.

The technique, which reanimates features using fairly simple software, is time-consuming but deemed one of the more dangerous forms of false news.

“Our idea was to 'inoculate' staff to pick these videos so that in future they would be able to detect important deep fakes,” said Baker.

While Deep Fake video is approaching, Michael Hayden of the Southern Poverty Law Centre painted an equally disturbing picture of the spread of hateful and false information. In a short but emotional and confrontational talk, he pointed out that as far as he was concerned, one of the more important bits of data released in the last few decades had gone largely unnoticed by mainstream journalists.

This, he said, was the first reports in publications beginning in the early 21st Century, particularly around the years 2001 and 2002, of census data showing that white people would become a minority in the USA within a few decades.

“This became the central recruitment point for hate groups and will be remembered as one of the most important bits of disinformation of the 20th Century,” he said.

Hayden went on to describe the pattern of behaviour of online hate speech and the link to violent mass shootings, many involving white males. Among these, he discussed the massacre of 50 Muslims in a New Zealand mosque by an Australian man armed with an AR-15 semi-automatic rifle.

“This data is at the core of the Alt-Right belief system and they use the census to reinforce their belief that Jews are behind some conspiracy,” said Hayden. He added that “people like Carlson and Steven Miller use the same narrative.” Tucker Carlson is a far-right commentator on Fox News, while Miller is an American far-right political activist who serves as a senior advisor for policy for President Donald Trump.

“They’re part of the group that shout 'Jews will not replace us' as they march at meetings,” he said, before warning that “the Myth of forced demographic replacement is what drove Trump to power and led to Brexit”.

The Southern Poverty Law Centre has been put under a great deal of social media related pressure through online threats. The effect on Hayden is clear - at times he seemed highly emotional. Being on the receiving end of daily hate mail and threats of violence have taken their toll on journalists.

A number of sessions at the conference have featured either a direct response to the threats, or alluded to them.

Hayden stressed that the fundamental problem was the use of data that was accurate to correlate with the extremist views and reinforce their conspiracy narrative.

“I asked a prominent white nationalist why he writes what he knows are lies and he said he’s doing it because it’s the best method,” said Hayden.

The challenge for international organisations is to deal with disinformation, which is particularly effective in regions where citizens have not had a history of free information flow.

One example of this is evident in the case of Shalini Joshi, an editor and specialist in verification at PROTO and the International Center for Journalists’ (ICFJ) partner for TruthBuzz in India. Her premise is based on a simple truth - it's now more important to explain how basic journalism works to the general population, rather than just discuss false news.

“There are many strategies that have the potential to be tried and tested in other areas,” she said. On example she provided was of a photograph that was shared before the 2019 elections in India featuring the headline: 'Trucks Loaded with Electronic Voting Machines are being sent around the state'.

“It happened in India but it could have been any other area where there is a growing concern about the vulnerability of voting machines, for example those used in US elections,” said Joshi.

False news on Indian Election

In a Vishvas News investigation, the claim that vote machines were being swapped was proven to be false, but tens of thousands of Indians had shared these images on Facebook and WhatsAppp. Source:

A more complex type of disinformation campaign with longer text messages has begun in India, and these campaigns are harder to debunk.

“There are also lengthy text messages, more visual content, memes, screenshots, and some of the opinions feature truths along with lies, so the misinformation makes it hard to verify.”

And encryption is making this worse, rather than better. Joshi pointed out that the closed messaging applications like WhatsApp mean that some facts cannot be verified easily unless the material is shared in open groups – or other platforms.

“End-to-end encryption means its becomes hard for fact checkers to figure out what’s going on, what’s trending and so on, there are no analytics offered by WhatsApp to help,” she explained.


There are solutions to this - but these will take time and effort. Joshi, for example, has been working with local authorities who are now being followed in their droves by first generation internet users. This was not the case previously. Joshi added that police stations are now using Facebook posts to fight disinformation campaigns, while her organisation was making use of local languages to offset the fake information.

She believes by making the material local, particularly in the specific style of language, the effect of disinformation is reduced. “It’s by engaging with people on their platforms, it’s very important to do this, with local people using listening posts on closed groups to help tip us off,” she said.

“Checkpoint in India created a database of rumours and false information and launched a tipping service where citizens could report what looked like rumours. They received over 150 000 messages, most of which were false news,” said Joshi.

“But we also need to watch Instagram, Snapchat and other apps. Those who don’t think these are important need to think again and build a collaborative process with local authorities,” she warned.

According to the panel members, Brazil faces even more of a challenge as its president employs Trumpian tactics to warp information. Brazilian fact-checking expert, Tai Nalon, said that verified information is hard to secure. She is the cofounder and director of Aos Fatos's, an organising regarded as Brazil's leading fact-checking platform, launched in July 2015.

“Brazil has a difficult task. It is a continental country where literary and access to internet connections are poor, but where we have some of the largest usage rates of social media.”

The country accounts for 10% of total time spent on social media globally, putting it in second place just behind the United States. It is no surprise that both countries are led by a populist president who uses social media as their main communication vehicle, bypassing traditional media outlets.

WhatsApp is the most popular social platform with 91% usage among internet users, followed by Facebook (86%) then Instagram (60%) and Facebook Messenger (59%).*

*Figures according to as of August 8 2019

Brazil social media use August 2019 - Source

“Apps aren’t developed for news, since 2016 smartphones overtook computers and now use has escalated,” said Nalon.

“At the same time, 87% people say they get their news from the internet, and by internet they mean social media platforms, while only 48% say they trust this news.”

Brazil has a polarised electorate, similar to that of the United States, where social media has worsened the political silos. This deep-seated dislike is escalated and exacerbated by digital platforms.

“There are low levels of media literacy,” she explained. “People feel that the zero-rated costs like for Facebook mean that people only get their information in these places. They don’t buy news online, they get social media news for free.”

Like India, people are sharing memes and pictures and find that small connections are more engaging, along with text messages sent by those they consider trusted friends.