Frances Haugen: ‘We are worse off today than when I leaked the Facebook documents’

In September 2021, The Wall Street Journal published the Facebook Files, a series of reports based on internal documents from the tech company that, among other things, showed its executives were aware of the harms Instagram and Facebook were causing young people. It was a bombshell. It triggered the biggest reputational crisis for Mark Zuckerberg’s company, which weeks later rebranded as Meta. The person behind it was engineer Frances Haugen, 42, who left her post at Facebook carrying 21,000 internal documents. The U.S. Senate summoned her to testify, and investigations were opened into her revelations.

The body of evidence provided by Haugen helped many parents connect the dots. Thousands of families with teenagers heavily involved in social media — who had suffered mental health issues, eating disorders, or even suicide — filed lawsuits against the company. Many of these cases became part of a class-action suit filed in 2023 by thousands of individuals and dozens of educational institutions against several social media platforms. That same year, attorneys general from 41 states sued Meta for harming children with its products and failing to disclose those risks.

The legal tsunami has already begun to yield results. Two months ago, a New Mexico jury ruled that Meta is guilty of misleading consumers about the safety of its platforms and putting minors at risk. In Los Angeles, Meta and YouTube (owned by Google) lost a case that found them liable for fostering addiction among minors.

Haugen now lives in Puerto Rico (she understands Spanish but doesn’t speak it). Her years at Google, Pinterest, and Facebook are behind her: she has founded an NGO, Beyond the Screen, aimed at making social media more transparent. She spoke exclusively to EL PAÍS in Barcelona after taking part as a keynote speaker at the first International Meeting on Digital Rights.

Question. It’s been five years since the leak. How would you describe Meta’s evolution since then?

Answer. During the first two years, it seemed they had gotten the message. A month after the leak, they announced they would double down on safety. And they did. They hired a lot of people and built a lot of systems. But then Elon Musk came along, bought X, and fired its whole safety and content-moderation team. He showed that there weren’t consequences for not investing in safety.

In a very short time, the industry as a whole made a radical turn. Company after company followed Elon’s lead. So I would say we may be worse off today than when I leaked the documents. We have not fully used the powers of tools like the Digital Services Act [DSA, the EU regulation to create a safer, more transparent online environment] or other systems that would allow us to put pressure on these platforms.

Q. The documents you leaked started a major legal process against Meta that has united families, educational institutions and state attorneys general, both Democrats and Republicans. How was it possible to generate so much consensus?

A. In the United States, we take for granted the idea that almost every important issue becomes this grinding and intractable fight where we can’t agree on how to move forward. Social media is one of those issues where left or right, people see the damage that’s happening to kids. And while we may not all be on the same page about how to intervene, there is consensus that teenagers should not be online at 2 a.m. and that, unless there is a plan for how phones are used in schools, we will keep suffering their distracting power. People are not willing to keep accepting those costs.

Q. What impact do you think the recent rulings in New Mexico and Los Angeles will have?

A. The New Mexico case is different from the others because it focused on child exploitation: child trafficking, sharing of sexual images of minors and child sexual abuse. I am incredibly heartened at how quickly the jury returned their verdict. Normally, in cases like this, if a decision comes in 24 or 48 hours, it tends to favor the defense. But there was so much evidence against Meta and it was so vivid that the jury decided immediately. I hope we see more states conclude that if New Mexico can win so decisively, they can too. It’s easy to view the New Mexico case as only involving $325 million [the size of the penalty]. That state has just 2.1 million residents, but if you scale those damages up to the size of the U.S., you’re talking about roughly $55 billion. And that’s only one case. If more states decide they will not accept the costs of child exploitation, teenage disorders or self-harm, that starts adding up to the kinds of damages that forces are reckoning.

Q. What will be the next step in this legal process?

A. We expect many more documents to come to light this summer as evidence in the California trial [the jurisdiction where the main lawsuits are filed]. We will also see the start of the federal phase of the class actions brought by families, individuals and school districts. So we’ll see what the different puzzle pieces are playing out over the next few months. That will be the next big front in this legal battle.

Q. How else should the harmful effects of social media be addressed?

A. Historically, when there are class actions on weighty issues that seem likely to succeed, Congress gets dislodged. Even if we have well-crafted laws ready to be enacted, they can languish because of the deadlock, as happened with tobacco. It was only after 46 states won that anti-tobacco laws started to be seriously discussed.

Q. Both in the U.S. and in Europe, there is growing concern about the effects of social media platforms on children, but it seems artificial intelligence is not part of that conversation. Do you think we will see a similar social reaction?

A. We have very little concept of what it means to be 14 today. It has become normal for teenagers to have AI friends. Character.ai [a site offering personalized conversational agents] brags that the median amount of time people spend with their avatars is two hours a day. Digital friends are not really your friends. Their enablers. They’re your sycophants. They might hype you up, but they’re designed to keep you on that system and isolated. They’re not designed to have you flourish.

In the U.S., we are starting to see lawsuits over self-harm and over children who died as a result of negligence and lack of oversight of these AI friends. If we don’t extend the conversation from social networks to digital friends, we will see many of the same problems repeat.

Q. Do you think class actions will also be brought over AI’s effects on children?

A. AI companies have faced lawsuits much earlier than platforms, comparatively. OpenAI, for example, only released ChatGPT three years ago and has already begun to face wrongful-death lawsuits [there are several documented cases of people allegedly driven to suicide by the chatbot]. Instagram went close to 15 years before they started facing serious legal consequences. During the first 10 years of social media, from 2004 to 2014, we thought it was fun and positive for the world. Companies have realized that if they don’t engage with the public more, they’re not going to fully blossom.

Q. What was the day after leaking the Facebook files like? Were you able to find work?

A. If you’re in a position where you could become a whistleblower, you’ve likely acquired enough life experience to do something else. When I worked at big tech companies, I didn’t realize they tried to make employees fear not being able to survive outside them. One thing that allowed me to take the step was YouTube’s algorithm recommending videos on how to start a small business. That’s how you get into this field.

I’ve gotten job offers back in corporate America, but I am pursuing other projects that I’m more passionate about. It will be interesting to see how Generation Z reacts in the next 10 years — they know they’re disposable and they don’t expect to stay long at one company. If you don’t have confidence that, even doing good work, you will remain at the company, then you have fewer incentives to keep corporate secrets.

Q. Right now there may be people thinking about leaking documents. What would you tell them?

A. Leaking is how critical information gets out, and you don’t need to become a whistleblower to give it to a journalist. Lots of people leak documents. Arturo [Béjar, also a former Meta employee] or I may be very visible, but for every Frances Haugen there are 100 people who make sure the right document gets into the right person’s hands. It’s important to know the proper ways to get a really important document out of a company. I, for example, took photos of my screen because I had a sneaking suspicion I was being watched by my employer.

Q. How do you imagine social media in 10 years?

A. One thing that has surprised me most in recent years is Blue Fever, which was among the top 10 downloaded apps in 2023. It’s a social network mainly used by Generation Z, and it’s different because it’s anonymous and focuses on feelings and experiences. It’s designed to be a safe space. We tend to assume that when people are anonymous, they behave badly. It opened my eyes to the idea that there are many creative people experimenting and that young people are eager to stop using Instagram and Facebook. I have a feeling that current youth movements will, within the next five years, find the key that makes platform choice a real option — so you decide with your friends where you want to be instead of being dragged back to the usual ones. The fact that many 21- and 22-year-olds don’t want to spend a decade doing what I did will create the critical mass that gives us a larger variety of options.

Sign up for our weekly newsletter to get more English-language news coverage from EL PAÍS USA Edition