Collaborative Journalism Summit: Projects to fight disinformation in elections popping up all over the world
Grégoire Lemarchand shared how CrossCheck worked in France and how its lessons could be applied in the U.S. this year
In his keynote presentation at the 2018 Collaborative Journalism Summit, Grégoire Lemarchand, the deputy editor in chief and head of social networks at Agence France-Presse, spoke about how American journalists might use lessons learned from CrossCheck, a collaborative verification project, in preparation for this year’s midterm elections.
CrossCheck, a collaboration between 37 partners across France and the United Kingdom, focused on covering “false, misleading, and confusing claims that circulated online” in the 10 weeks before the 2017 French presidential election.
“Participants were under a common sense of public service,” Lemarchand said. In the weeks leading up to the election, he said, the French public’s trust in the media was very low — dangerously low. “There was really a global feeling that the threat of this disinformation is so big and serious that we have to work together. If we allow this information to spiral out of control, we will be left crying, as the public will no longer know what is true. If people cannot trust, then democracy can’t work.”
CrossCheck, which was launched by Jenni Sargent, the managing director of First Draft, wasn’t just limited to newsrooms — journalists were joined by teams at universities, nonprofits and tech companies. Google News Lab in Paris and Facebook gave financial and technical support to the project.
The project began as an experiment, and Lemarchand said several partners were unsure about its eventual success. Several questions floated in the air. Would newsrooms be able to support additional work? Would journalists be able to work together? Would readers trust a new brand?
- Read the final research report on CrossCheck “The Impact of CrossCheck on Journalists and the Audience.”
Collaboration hailed as one of best outcomes of CrossCheck project
“Collaboration […] was certainly one of our most positive achievements,” Lemarchand said. “Decisions were taken collectively. We never had strong disagreements, only debates, with everyone benefiting from the various expertise available.”
The project’s goal, simply put, was to help French voters make informed decisions. But in a world of rumor, hoaxes, and outright lies, feeling confident as a voter isn’t as easy as you might expect.
By the end of the project, the team had published 67 articles in French and English (from Feb. 27 to May 5, 2017) debunking disinformation. The project tackled more than 600 questions from the public. Several partners published CrossCheck’s work on their own sites and platforms, which extended its reach.
Other metrics of the debunks that were published included:
- 590,000 page views and 336,000 visitors.
- 1,200,000 video views.
- 172,000 followers on Facebook, with 347,800 engagements.
- 4,900 followers on Twitter, with 1,900 mentions and 6,100 retweets.
After a quick ramp-up, CrossCheck establishes detailed workflow
The project grew fast: it launched in only seven weeks. The first meeting, in January 2017, convened 40 journalists in Paris. A two-day boot camp came a month and a half later. Then, on Feb. 27, 2017, the project officially launched at Agence France-Presse’s headquarters in Paris.
Once launched, the team began this workflow:
- Newsrooms and journalism students monitored social media with such tools as Hearken, CrowdTangle and News Whip to identify cases of disinformation. They also took in questions from the public.
- Over Slack, the teams collectively decided which issues were worth debunking, and worked through a thorough fact-checking and verification process.
- Journalists added their media company’s logo to the story as an endorsement. At least two endorsements were needed for a story to be published.
- The journalists wrote up an explanatory section identifying the sources of misinformation, with links to source material and a unique visual icon. The icons gave readers an at-a-glance impression of the story: Misleading, Manufactured, Manipulated, Misreported, Satire and Misattributed.
- The story was sent for review to an editor at Agence France-Presse.
- Finally, the story was published and shared on social media.
Lemarchand admitted that the cross-checking process is slower than traditional reporting. “There was sometimes a bit of frustration, because the publishing process was a bit slow,” he said. “We have to wait for the media endorsements, for example.”
But, he said, the payoff is big: high quality journalism that readers can trust. Explaining how a story was verified and fact-checked not only “increased trust in the article, but also helps readers learn they could do this work themselves,” he said.
Lessons include higher threshold of involvement for partners, better social reach
Lemarchand noted a few key questions for future verification projects:
- What topics should newsrooms focus on verifying next? CrossCheck focused on politics in the 10-week lead-up to the French presidential election. “But every time you’re talking about politics,” Lemarchand said, “a lot of people will always see bias or political orientation. So as we are seeking a future collaboration, we all think that we should go far beyond the politics, and work on other topics like health, education or the environment.”
- How should newsrooms decide what disinformation to report on/debunk, and what to ignore? “There is the threat of giving more exposure to rumors. We all know that,” Lemarchand said. “This is maybe the worst nightmare for debunkers: to give additional oxygen to false information, and to move it out of niche online communities to wider audiences.”
Lemarchand also noted a few opportunities for improvement:
- Increase involvement from every news partner: The project suffered when some partners didn’t meet their entire commitment. Ideally there would be at least one fully dedicated person for each newsroom, Lemarchand said, but he acknowledged that budgets and other constraints might make this difficult.
- Extend social reach: Lemarchand said his “biggest frustration” was that “we suffered from a lack of visibility on social networks, at least at the beginning.” But the project solved that problem by increasing ad spending on Facebook. They also began producing short explanatory videos, which worked so well that Lemarchand said that he wished the team had produced more.
- Evangelize and improve perception of project: Lemarchand admitted that many newsrooms are reluctant to devote time and resources to fact checking and debunking, as other work (e.g. reporting, editing) is top priority. “There are still some people in my home newsroom who think that CrossCheck was something like a gadget…[not] serious journalism. It’s a pity. It’s a shame,” he said.
Watch his presentation below. You’ll need to scrub through to the 17:00 mark, which is where his keynote picks up.
Jeanette Beebe is a freelance journalist.
About the Center for Cooperative Media: The Center is a grant-funded program of the School of Communication and Media at Montclair State University. The Center is supported with funding from the John S. and James L. Knight Foundation, the Geraldine R. Dodge Foundation and Democracy Fund. Its mission is to grow and strengthen local journalism, and in doing so serve New Jersey residents. For more information, visit CenterforCooperativeMedia.org.