“Social media platforms are shutting down their APIs. They're making it more difficult for academics to do this sort of work, which I think is really worrisome. I really hope that there's some action in terms of thinking through: if we want to understand what's happening on these platforms, we have to have some mechanism that propels the platforms to do it, because relying on their goodwill every once in a while probably isn't a sufficient strategy.”
The Lawfare Podcast: Patreon Edition · US Politics · October 2026
Context
Asked about the policy implications of the research, Stroud notes the study happened only because Meta allowed it and that outside data access has worsened. She argues for a mechanism that compels platforms to share data.
Transcript
…because the platforms, because a platform decided it's great that it happened and it's great that this platform decided it would happen. But yeah, as a citizen, that's not the world I want to live in. I don't want it to be only up to Mark Zuckerberg whether this type of research takes place.
…you know, what this says in terms of setting policy for how research on platforms is conducted. And I'd certainly be curious if either of you have any additional thoughts about that. There's also the question of, you know, substantive policy implications, if anything flows out of your findings.
Yeah, I mean, just as a starting point, I think it's absolutely the case that this research was done by the will of the platform. And credit to Meta for doing it, but the data access is so poor. And if we look back even five years, our data access has become worse. Social media platforms are shutting down their APIs. They're making it more difficult for academics to do this sort of work, which I think is really worrisome. I really hope that there's some action in terms of thinking through: if we want to have these data, if we want to understand what's happening on these platforms, we have to have some mechanism that propels the platforms to do it because relying on their goodwill every once in a while probably isn't a sufficient strategy in order to get data access. So I think that's a really critical part of this research is demonstrating the sort of things that we can learn from the platforms, in this case, in partnership with them, but this is just one case on one social media platform. And so I think the opportunity to really dive into what sort of observational data do we want to know? Do we want to know how many instances of information from untrustworthy sources people are seeing on platforms? It seems like that sort of even…
that that change does a lot to affect what people see in their feed. And some of it We probably would say, oh, yeah, that's a great thing. Switching to chronological feed reduced exposure to uncivil content on Facebook. Maybe we'd say that's a good thing. Maybe not. It's worth debating. But there…
think that it exposes the complexity of policy solutions that dictate how algorithms have to operate. It's just not so simple in terms of one switch to an algorithm has all these changes in terms of what content has surfaced and what people see on the platforms. So I hope that the implication of…
beginning that data access was poor, I just want to clarify that didn't mean data access was poor in the context of this study. Our data access
More from The Lawfare Podcast: Patreon Edition
