Laura Zommer, co-founder of Factchequeado: “Collaboration is our best option to fight misinformation”

Haga clic aquí para leer este artículo en español.

Coming from a family of Cuban immigrants in Miami, I have the stereotypical aunt and uncle who share viral links in our group chat on WhatsApp about “how to cure a disease with lemon” or “how the Northern Lights were actually caused by humans and not a solar storm.”

Sometimes the topics are mostly innocent, but other times they have real effects on the health and well-being of communities. For more than 60 million Hispanics in the United States, this may be a common scene in WhatsApp group chats, especially those where most participants only speak Spanish.

In the fast-changing and uncertain world of generative AI that we live in today, there is a small but mighty team of journalists and technologists who have been fighting misinformation in the U.S. for over two years, focusing on Hispanic communities. During a Zoom video meeting earlier this year, I had the pleasure to speak with Laura Zommer, a journalist and lawyer born in Argentina who is the co-founder and CEO of Factchequeado, a nonpartisan, nonprofit initiative that seeks to fill the void of reliable information in Spanish through innovation and alliances in the United States to effectively combat misinformation.

Zommer first started the battle against fake content around 2010 in Argentina, where she led Chequeado, one of the first fact checking initiatives in the Global South, and created LatamChequea, a regional network of fact checkers that currently partners with 41 media outlets from 19 countries. It was after realizing that they were facing the same challenges as Maldita, another fact-checking organization from Spain, that she and Clara Jiménez-Cruz, CEO of Maldita, decided to join forces to help close the gaps in the information needs of the Spanish-speaking population in the United States.

“We joined and launched Factchequeado in 2022 because we were seeing that the efforts against misinformation in the U.S. in Spanish were small, scattered and not necessarily taking into account some important lessons that our organizations — Chequeado, with more than 12 years, and Maldita, with more than six — already learned,” Zommer said.

In April 2022, Factchequeado started as a hub that produced and distributed free quality content that media partners could publish in Spanish, focusing on viral widespread disinformation topics. Today, they have expanded their network to over 100 partners who access not only the content but other tools like Chequeabot (developed by Chequeado in 2017), which helps with fact-checking and reformatting content; or a Whatsapp chatbot (developed by Botalite, a partner company of Maldita) that offers verified content and a direct channel to submit fact-checking requests.

I tried the chatbot on my personal WhatsApp, and as a web applications engineer, I was also curious to hear how Zommer and her team built this product and the strategies behind the Factchequeado alliance with U.S. media and civil society organizations.

Marita: What inspired you to start Factchequeado?

Laura: We know from different surveys that Hispanics or Latinos who live in the United States trust the mainstream media little, because they do not feel sufficiently included in their agendas. Based also on the experience we had in Argentina and Spain, it seemed central to us to listen more and help build a community that would make decisions on our journalistic agenda, based on what people and colleagues were asking us to cover.

In the past, there were few media in the United States paying attention to the information consumption of these communities. Typically, they consume more information content on social networks, in a higher percentage than the average American, and have an intensive use of WhatsApp, not only to communicate with families and friends, but also to inform themselves. We launched Factchequeado because we wanted to bring to the table some learnings that we had, especially related to paying attention to the audience to build the agenda, being in the spaces and channels where these groups get their information, and producing content in audio, video or in graphic files — the formats that these groups use to consume information.

Disinformants in general — those who earn money or power with disinformation — work in a structured way. In Latin America, and now in the U.S., we started to form those structures too, but to counteract misinformation. Today, in the United States and any other part of the world, collaborating is our best option to fight misinformation and have a real impact. We don’t ask ourselves if we can or cannot collaborate. What we ask ourselves is how we can collaborate more effectively, how we can design the best possible workflow so that this can extend over time.

M: Audience listening is at the center of the work you do. What other factors are crucial to your project?

L: It is those people who consume more disinformation and do not necessarily have a varied information diet who we are trying to reach. Another important step for us is helping to educate using tools that detect or combat disinformation, and tools to identify the use of artificial intelligence in images or videos.

Recently, our then editor-in-chief Tamoa Calzadilla collaborated with the Maldita Educa and Chequeado Educación teams to publish a bilingual guide for journalists covering Latino communities with 17 tools to combat misinformation as part of the institutional fellowship from the Reynolds Journalism Institute. At Factchequeado, we share advice and knowledge not only for journalists, but also directly with the public, with educational content on how to verify information and find reliable sources. In fact, this year, with the support of PEN America, we launched Fact-Desafío, a course specially designed to teach Media Literacy on WhatsApp.

M: The project helps with content production and distribution, as well as using tech to facilitate those processes. What have been the biggest challenges for your team during these first years?

L: In general, even in big media, the teams that work against misinformation in Spanish are small. The same thing happens with big tech companies. The treatment they give to misinformation in languages ​​other than English is unequal. And that is what we experienced in the last years: the same content, sometimes the exact same video, circulates in both English and Spanish but does not have the same treatment. Part of what we are doing with Factchequeado, in strategic policy terms, is to draw attention to this problem and inequality.

The size of the Hispanic community in the United States is much larger than the investment that the United States, whether the public sector or the private sector, is making in favor of quality information in Spanish.

Another challenge is that we need to be aware that disinformation itself, our experience shows us, has both a global component and a hyperlocal component.

M: That reminds me of the term that has been around for many years, “glocal,” which gets even more relevant in the current news ecosystem. Do you have any examples of how this works in the context of misinformation?

L: In the pandemic we saw many examples of that. In Mexico there was a narrative that you could protect yourself from Covid “by eating avocado,” in France “by eating cheeses,” etc. Well, it was the same narrative, but with a characteristic of the country or the community where they appeared. We believe that to be more effective in counteracting disinformation in Spanish in the United States, we need to have a global and also a local perspective. That is why we have our model of over 100 allies in 23 states and Puerto Rico, and we keep growing. Those allies are the ones who listen, know and work in the different cities with their communities, and in some way they are our ears in those places saying what information void is worth filling.

In the U.S., the local context is very important. For example, it is a common mistake to talk about “Latinos in the United States.” But Cubans in Miami are different from Mexicans in Texas, or Puerto Ricans or Salvadorans who have just arrived in New York. Understanding this diversity and this complexity is also key.

In general, we know very little about Spanish misinformation in the United States. We need to increase awareness — more people need to know that misinformation is a problem that can cause real effects, it can cause risk of life, effects on health, effects on democracy, effects on the financial system. And those people from the Hispanic communities, as they become more aware, they also understand that they can contribute to making the problem worse or be part of the solution.

M: Could you share some of the most effective strategies you have used to fight misinformation?

L: A recommendation that we make to any media, whether in Spanish or English covering Latin or Spanish communities, is to look at the opportunities to make information already available in English more accessible. Even when a tornado alert arrives in English in this country, there are people who will be excluded from that because they do not understand it. A great recommendation to all of us is to identify the information gaps in advance and try to fill them proactively. When the elections come, the media realize that they have to work on the Latino vote, but by then it’s too late. What we have to do is identify those gaps in advance.

M: In a year of elections, what are the main trends that news organizations need to keep an eye on?

L: This year, we will have elections in many countries, and it will be the first time we have elections during the boom of artificial intelligence in the U.S.

Based on previous years, there are some narratives that we are paying attention to in advance. We launched a series called Electoral Explainer, which outlines 10 disinformation narratives linked to the elections that were already present in previous years. We do it in multiformat, in part because electoral misinformation often follows the same patterns no matter when or where. “There will be fraud, or there was fraud” — that same narrative occurs in the United States, but it also appeared in Brazil and other countries.

So what do we have to do? Explain as much as we can how the process works, how secure or not those processes are. Newsrooms, too, can explain the rules so that those who want to put more shade on the process do not take advantage of gaps or misunderstandings.

M: We mentioned AI briefly, but I would love to know, in your opinion, how AI helps or damages the mission of fact checkers to fight misinformation.

L: One of the main goals of fighting misinformation is to deliver accurate information quickly and in the formats the audience consumes them. Artificial intelligence can help us better monitor  trends and act faster. Last year, with the support of NYU and the Chequeado developer team,  we made a tool to improve misinformation monitoring in Spanish in the United States, and now that tool is already a viable product and we are starting to test it with allied organizations. AI can help us transform text into multimedia formats faster as well. So if I have a small team of editors and journalists, I will have more opportunities to produce more content in less time, with more chances of having a positive impact.

What scares us? What worries us? Deepfakes were more difficult to make a few years ago. Now the cost has dropped, and tools are very available.

Recently, content circulated with the voice of journalist Jorge Ramos, who in the Hispanic community is very recognized, supposedly announcing help from the government for immigrants and foreigners. It was a deepfake generated with AI: Jorge Ramos had not participated in that, and the government had not launched the help. It came from someone who probably wanted to make money with the clicks they received from these notes. The same thing happens with images that can make us believe that a candidate is hugging an opponent, etc.

But we also have to be attentive to what is called cheapfakes, because due to how social networks work, a Photoshop of an image or a photo taken out of context can spread misinformation.

The key, I believe, is to choose where to concentrate the limited resources we have, build networks to be more efficient and have more impact, develop technology that helps us increase our reach at a bigger scale, and aim to educate and train audiences in critical thinking.

Reflect on these ideas in your news organization:

• What is the process to fact-check information in your newsroom?
• What segment of your audience is more susceptible to consuming misinformation? How can you target them better with fact-checked information?
• What tools can you add to that process? Is there any automation you could put in place that helps with distribution?
• In the news ecosystem of your community, what alliances can you establish to work towards fighting against misinformation?

If you have any ideas, suggestions or experiences around fighting misinformation that you would like to share with other newsrooms, you can email me at marita.perezdiaz@pressinstitute.org.