RTVE’s Verification Team in the corona virus crisis
By Iris López de Solís, Documentalist at RTVE – published on 22 April 2020
The ‘Value Use and Copyright Commission’ (VUC) of FIAT/IFTA has decided to publish a series of interviews and articles about use of audiovisual archive materials. This article explains how RTVE’s fact checking team fights several forms of misinformation circulating in Spain and related to the corona virus outbreak. Archival and information management skills are a crucial contribution to the work of this team.
Mid-February, the World Health Organization (WHO) announced that the new coronavirus pandemic was accompanied by an ‘infodemic’ of misinformation. On March 14th, Spanish Prime Minister Pedro Sanchez declared the state of alarm in Spain. Since that day, the RTVE pop-up verification team has been fighting against misinformation spreading in the social networks about the health crisis.
The RTVE verification team, pushed by RTVE´s digital area, worked previously during the two general election campaigns of 2019. The team took part in the Comprobado, a collaborative journalism project coordinated by First Draft , whose main mission is to empower societies with accurate information in critical moments.
Similar to back then, RTVE has now again set up a multidisciplinary team with people who not only have different professional profiles but also work in different departments and areas. Currently, the team is formed by:
- three journalists from the RTVE Audiovisual Innovation Lab
- an RTVE website journalist
- a journalist specialized in international verification
- three TVE documentalists
A recent survey conducted by the Reuters Institute shows that rather than being completely fabricated, much of the misinformation on Coronavirus involves various forms of reconfiguration when existing and often true information is spun, twisted, recontextualised, or reworked. So the main objective of RTVE team is to debunk those half-truths and hoaxes that circulate both social networks as through instant messaging services.
Following the analysis of the First Draft, during this month of work the team identified different subjects of misinformation around the Coronavirus:
- Conspiracy theories about where COVID-19 came from, e.g. the virus was created in a laboratory by Microsoft founder Bill Gates to lower the population number or it is a weapon used by the United States against China.
- How the virus spreads and can be transmitted. Some of them are curious, like the risk involved in going out on the balcony to clap.
- Symptoms and treatments: these are by far the most common forms of disinformation and also the most dangerous. For example, recommendations for taking a different illegal drug to prevent or treat Coronavirus. Nowadays, scammers are using the health crisis to sell false testing tests, cures and treatments.
- The response of the Spanish Government and citizens: an anonymous audio alerting of a coming state of exception and the closure of supermarkets was quickly spreading through WhatsApp, causing great social alarm among Spanish citizens. As a result, police had to officially deny it in the daily presser on Coronavirus. Also, a video showing pictures of a crowd outside a supermarket in Germany in 2011 during the sales was reframed as shot in Spain.
The team also noticed that the amount of hoaxes and misinformation is much bigger than it was in the last electoral elections and closed messaging applications are now more popular for sharing rumours and half-truths because they can be more easily spread and, in most cases, they can´t be tracked. Both in WhatsApp and Telegram visual contents and memes are a compelling vehicle for misinformation.
A screenshot from the verification site rtve.es/noticias/coronavirus-covid-19/bulos-fake-news
Regarding its verifications and publications, the team is prioritising different contents:
- Issues that affect citizens directly. Priority is always given to those issues related to pseudo-medical recommendations to treat the coronavirus.
- Issues that cause extra alarm and threaten calm.
- Issues that threaten cybersecurity.
- Subjects that threaten the dignity of a person, like words put in somebody´s mouth when that person didn’t utter them. For example, audios pretending to be from a former Spanish Minister or a famous judge.
But also the virality of the messages is taken into account and the tipping point to avoid adding fuel to the fire and make greater hoax out of what otherwise might have faded away. So, following the First Draft´ advice, some useful questions help us determine that “tipping point”:
- How much engagement is the content getting?
- Is the content moving from one community to another?
- Is the content moving across platforms?
- Did an influencer share it? According to the Reuters Institute high-level politicians, celebrities, or other prominent public figures produce or spread 20% of the misinformation. But that misinformation attracts a large majority of all social media engagements (69%).
- Are other journalist and news media writing about it?
Every day the verification team monitors different social networks. For example, Crowdtangle is used to detect hoaxes spreading on Facebook and Instagram. The team also follows several hashtags and keywords popular on Twitter with Tweetdeck. Next to that, it has been able to monitor some closed Facebook groups and Telegram groups supporting different political parties. Moreover, it receives messages, videos and audio files sent by citizens to a WhatsApp number. And team member’s friends, relatives and colleagues play a key role to get these contents to the team.
All the verifications are based on fact-checking techniques, including checking official statistical data, forensic analyses of images, reverse search of audiovisual contents or identification of false social media accounts. When needed, institutions or people mentioned or implicated in the information are contacted.
The verification team publishes its conclusions mainly on their website and social networks. At least one article comes out daily and contributions are added constantly, minute by minute, on the web. Besides, the team has the opportunity to explain investigations in two sections on two radio stations: Radio 5 (nationwide) and Radio 4 (in Catalonia region). An extra editor from Catalonia collaborates with the team by producing her own scripts.
“… the amount of hoaxes and misinformation is much bigger than it was in the last electoral elections and closed messaging applications are now more popular for sharing rumours and half-truths …”
The verification team also has a weekly section on TVE targeted to citizens with little technological skills and who are afraid of cybercrime. It receives requests from the TVE newsroom to verify pictures and videos from social networks. Sometimes, journalists ask for audiovisual hoaxes when they become part of a newsworthy event. So, in these cases, it plays the role of an audiovisual archive.
Apart from the work on the web and radio, the team answers citizens’ questions and hoaxes one by one. So in this previous month, more than 200 requests were resolved. Moreover, with all this information a questionnaire is fed to make an audio product for Alexa, the virtual assistant AI technology developed by Amazon.
The RTVE team is collaborating with EBU’s Coronavirus fact check initiative as well. As a result, the team verified a fake video assuring that a RAI programme predicted the Coronavirus pandemic in 2015 and it was able to prove that a video spreading as being shot in a hospital in Spain was filmed originally in Guayaquil (Ecuador).
Based on their experiences, the verification squad considers the decision to set up a multidisciplinary team as a positive one, as each member contributes to the project with different knowledge and expertise. Regarding the role of documentalists, their work is useful for the team since they are used to looking for information and data not only on the web, of the official organisations or on social networks, but also in the RTVE archive and other archival collections. Even so, the team is giving online courses on verification for colleagues during the lockdown.
The team’s main goal for the future is not only to be a pop-up but a stable team, with new challenges and goals, taking part in international collaborative fact-checking projects and including AI in the verification workflow. For the time being, as a public service, the verification team is continuing its fight against those “bad actors” producing misinformation amid this infodemic and helping citizens to empower them against disinformation.