Since 2018, Professor Joan Donovan has been Director of Research at the Shorenstein Center at Harvard University’s Kennedy School. His job was to lead a team of researchers into disinformation and manipulation campaigns on networks and to obtain funding for their projects. Apparently everything was fine until Wall Street Journal In September 2021, he began publishing a series of articles containing internal Meta documents, leaked by former employee and “deepthroat” Frances Haugen.
Joan Donovan obtained a copy of the thousands of documents Haugen had removed from Meta. I wanted to organize them and publish them on a website called Facebook Archive at Harvard. Then the problems started. A meeting in October 2021 with his superior, Dean Douglas Elmendorf, and other figures, including a former Facebook manager, was a turning point: “At that meeting, when a Facebook executive became angry against me, something changed at the Kennedy school, where I later became a outsider and they put me on the defensive not only for explaining my research, but also for explaining my research specifically on Facebook,” Donovan told EL PAÍS via video conference.
At that meeting, Donovan said Haugen’s leaked documents were “the most important in the history of the Internet.” A few days later, Donovan received an email from Elmendorf to “discuss his research”: “Since that meeting, my time at Harvard has not been easy,” he says. In 2022, Donovan learned he would not continue in his position. In February 2023, it was released to the public and ended up being released in August. Last December, Donovan denounced Harvard because he believed the university had bowed to Meta’s wishes: “My story is a little complicated, but the way I like to explain it is that Elmendorf a very long relationship with different Meta leaders, in particular Sheryl Sandberg (Meta number 2 until June 2022). What we don’t know is to what extent Facebook or Meta pressured him to stop my searches,” he admits. Donovan’s contract as a teacher lasted until December 2024.
500 million in 15 years
The story of Donovan’s departure due to alleged Meta pressure on his dean reaches another milestone in December 2021: the largest gift in Harvard history. The Chan Zuckerberg Initiative, created by husband and wife Priscilla Chan and Mark Zuckerberg, will donate $500 million to the university over 15 years. “There is growing evidence that donors have undue influence over Harvard. My expulsion coincided with the largest gift in the university’s history. But the money will not come suddenly, but over 15 years. So, Facebook has already gained 15 years of prestige and influence on the largest university brand in the world,” says Donovan. For his research, Donovan had raised 12 million, which is a considerable sum and Harvard still has more than 3. “I hope that they will return it to the donors so that they can distribute it again”, specifies the researcher .
Meta has not commented on this information and the Chan Zuckerberg Initiative released a brief statement saying his donation had nothing to do with these events.
The text of Donovan’s complaint, replete with messages with Elmendorf and other teachers, compares the practices of Meta leaders, first, to those of “foreign intelligence services or organized criminal enterprises” in their stealthy influence efforts. And second, to the tobacco and oil industries: “If we look at the field of Internet studies and look at who is receiving funding for what type of research, Facebook is clearly giving millions of dollars to researchers in an effort to boost media coverage positive. » of their platforms. And then among the researchers who criticize them, there have been different cases in which Facebook has tried to influence publication or to get academics not to publish. That’s why it looks a lot like the playbook of the tobacco industry, big oil companies and pharmaceutical companies. In terms of using academia as a form of public relations for business.
This summer, a large group of renowned researchers published a series of articles in prestigious scientific journals Nature And Science in collaboration with Meta employees. Donovan considers this a dubious and short-term project: “I can’t call them truly scientific, because the researchers asked Facebook questions, and then the Facebook researchers gave them an interpretation of the data. independent They have never seen the data. In my opinion, when you do scientific research, you have to collect your own data,” he says. Additionally, Meta took advantage of the project to boast about the results in a biased manner, according to Donovan: “Facebook knew that most people were not going to read the full articles. “Meta released a statement saying the study showed there were no problems with their algorithms or radicalization in their products,” he adds.
Harvard University denies Donovan’s accusations, arguing that this research project required the supervision of a university professor (Donovan was hired). And they didn’t find it, according to a spokeswoman for the institution: “After these efforts failed, more than a year was given for the (Donovan) project to be completed. “Joan Donovan was not terminated and most members of the research team chose to remain at the school in new roles.”
Today, Donovan is a professor at Boston University, although Harvard is the intellectual owner of his research in those years. Donovan’s suit is aimed in part at getting his academic work back to continue working at his new center.
Donovan says he would have easily found a new academic responsible for his research. Harvard also states that the famous Facebook archive with Haugen documents is online, as Donovan had planned. The file is, however, very different from that projected by Donovan. The search engine allows you to search by keywords and the system simply returns the slides to where they appear. But these are loose captures whose context is difficult to discern. The investigation of Newspaper This required several months of work by a handful of journalists: “Harvard took the Facebook archive project from me. He posted it, but it’s almost useless. It’s difficult to navigate. My vision was to create a database that would enable international collaboration and understanding of these documents. Various governments around the world are trying to understand the negative influence of Meta on adolescents and young users. “Meta knew there were problems with Instagram and they didn’t do anything,” Donovan says.
Another important concern of Donovan is academic freedom. When the dean tells her the end of her project, he warns her that she is not protected by academic freedom: “There are approximately 6,000 researchers at Harvard who publish and who are in the same situation as me. And no one who publishes controversial research will do so if their university doesn’t protect them. “If I stayed at Harvard and posted these articles on Facebook and got sued, I was on my own,” Donovan says.
The researcher believes that her complaint, filed with Whistleblower Aid, an organization that helps people seeking to protect themselves against the disclosure of trade secrets and which has previously collaborated with Haugen, is just one example of what Meta is doing as much as it can: “My story is ‘This is just one small piece of this big puzzle in which we need to understand how this company and others shape our institutions so that there is no regulation and there is no clear investigation into the real harm these platforms are causing to society.”
Donovan’s experience in other cases leads him to be even more wary: “What I know about Meta’s PR strategies to try to kill certain stories is that they don’t start with the reporter to convince him that his story has no validity. They start at the top, they start with the board in the middle. So I wasn’t surprised when a week after the meeting I received an email from the dean repeating all the PR gibberish on Facebook,” he explains.
Meta is not the only one of these platforms that can, more or less inadvertently, cause social harm. For Donovan, the difference is what they do when they know their products have defects: “All of these technologies cause similar types of damage, but what the company does about it is as important as the design of the technology itself. And if the company hides or conceals the impact of its products on democracy or public health, then we need to dig deeper to better understand and explore it, because the company itself won’t do it,” adds Donovan.
Subscribe to continue reading
Read without limits