Threads, Meta’s new Twitter competitor, continues to grow at an astonishing rate: In the week since the app was released, it has already amassed over 100 million users.

But for all that Threads might offer in terms of reliability, the backing of a big tech company, and less chaos than Twitter under Elon Musk, there’s also a very real fear among academic researchers that vital visibility into how we interact online—which increasingly bleeds into the offline—is going to disappear.

That’s because Meta has historically been more reticent to share insights about its users with academic researchers than other platforms—particularly Twitter.

“Twitter was, by and large, the best dataset to understand how people interacted on social media sites, just because most of the posts were public to begin with and the API—before it was behind the paywall—was relatively accessible,” says Rebecca Williams, associate director for privacy and data at the American Civil Liberties Union. “It was the best dataset, not only to observe how people act on the internet for content moderation, but also just to do sociological research on how people felt about elections and things like that.”

Meta, however, has long been the opposite, including shrinking support for Crowdtangle, its fact-checking tool that can track the spread of posts on its suite of apps, in recent years.

“Meta properties have traditionally been closed off to researchers, especially after the Cambridge Analytica scandal,” says Jeremy Blackburn, an assistant professor at Binghamton University in New York. “Meta properties are also incredibly aggressive with respect to requiring logins to view content: Instagram will only let you view a handful of pics before asking you to login.” Case in point: access to Crowdtangle is mediated through a Facebook account.

Because of that, Blackburn says he’d be surprised if Threads ever became a replacement for Twitter in academic research. “There is no reason to believe that Meta will alter their policies, and whatever access they might give scientists under these policies is suspect at best.”

A Meta spokesperson tells Fast Company: “We’ve been sharing data that supports public interest and scientific research for the past 10 years and remain committed to working with researchers to continue this effort. We’re currently rolling out new data sharing tools for researchers and will assess the inclusion of Threads as part of these efforts while maintaining the privacy of people who use our platforms.”

Until Twitter owner Elon Musk started trying to cash in on API access earlier this year, charging researchers $42,000 a month to access tweets for academic purposes, Twitter had been a surprising bastion of transparency on social media. It was willing to let academics peek under the hood, embracing its role that Musk had identified when he launched his bid to buy the platform as the “de facto public square.”

Meta, researchers worry, could limit scrutiny of the public square. That would have significant negative effects on our understanding of society and how it’s evolving. We spend more and more of our time online, and so studying how we interact on those platforms is increasingly important. The data researchers uncover by looking at a platform from 20,000 feet can help glean insights into where they work well—and badly. Meta’s more controlled access to that data would “mean our legislative bodies have even less information to work with,” says Williams. “And our society has even less information to work with on how these platforms are working at scale.” Williams says she has previously spoken to members of the Facebook Oversight Board, an independent group designed to ensure the platform makes decisions that are proportionate for its user base, who said they would like to see the company be more open in its sharing of data about its platforms.

And yet despite that, there are reasons to be optimistic about not losing all visibility over the “de facto public square,” as Musk once called Twitter, and which Meta presumably is eager to replicate. “Irrespective of access to the API, researchers will always be able to obtain data from the platform,” says Savvas Zannettou, an assistant professor at Delft University of Technology, pointing out that it’s always possible to scrape such platforms for their data. “It will be a bit more work, but we can still audit and analyze online phenomena,” Zannettou says.

Another bright spot is the European Union’s Digital Services Act, passed late last year and coming into force in January 2024, which contains clauses allowing researchers to apply to obtain data direct from platforms to study potential harms that arise from the use of their services. While platforms may try to make it tricky to obtain that data, Zannettou believes they will eventually be compelled to hand it over because the risk of being fined or sanctioned by the EU will be too great.

Zannettou also thinks people power could play a part in helping make platforms more transparent. Users can, for example, download their data directly from the apps and then donate it to researchers. “Research will always continue, irrespective of data access provisions by the platforms,” he says.