Why TikTok, X, and Meta are accused of blocking EU researcher data access
European academics say TikTok, X, and Meta are not providing platform data they are legally required to share. According to the researchers, that gap is significantly hampering independent studies of social media's effects on public opinion, mental health, and the spread of disinformation. The companies themselves are not directly quoted in what has been reported here; the claims are attributed to researchers and a related complaint.
The obligation in question stems from European Union rules requiring large platforms to provide researchers with data access. Rules of this kind are meant to let academics and civil-society groups independently examine platform behavior — how algorithms surface content, for instance, or how quickly misinformation spreads.
According to the researchers, there is a wide gap between the access rules grant on paper and what actually gets delivered in practice. Data requests are said to be delayed, the data that is provided is said to be incomplete or delivered in unusable formats, or access conditions are said to be restrictive enough to make research effectively impossible. These claims have not been independently verified as part of this report.
It is worth stressing that no company's violation of the rules is presented here as an established fact. Words like "accused," "say," and "according to researchers" are used deliberately, because what is known rests on statements from researchers and the underlying complaint or report; the companies are not quoted with a response in what has been reported here.
This is not a new debate: platform transparency and researcher access have long sat at the center of discussions about social media's effect on society. Researchers argue that, absent independent outside scrutiny, statements platforms make based on their own internal data remain unverifiable.
The three companies are not in identical positions in this dispute; each has a different product and user base. But according to the researchers, the data-access problem is observed across all three, in similar or differing forms. No loaded language is used toward any of the three companies in this report; a neutral framing is maintained throughout.
A lack of data access can have concrete consequences: researchers may struggle to produce evidence-based findings on questions such as how algorithmic recommendations contribute to polarization, how quickly misinformation spreads, or the effects on younger users. That, in turn, directly affects how much the public and regulators can actually know about platforms' real-world impact.
How enforceable the regulatory framework proves to be is also part of this debate. Rules may exist on paper, but how they are enforced, and how allegations of violations get investigated, determines how effective the regulation is in practice. That is not unique to these three companies; it is a broader test case for platform regulation generally.
Going forward, it will be worth watching how the underlying complaint or report is handled and how regulators respond. How the companies respond to these claims will also be an important factor shaping how the public assesses the matter.
Taken together, the dispute raises a broader question about how far data-transparency rules can close the gap between intent and enforcement — not as a resolved outcome, but as part of an ongoing debate.
Read next
Pixel 11 price hike: Google's Shakil Barkat all but confirms it
Google's VP of Devices and Services, Shakil Barkat, all but confirmed in an interview with 9to5Google that the company's next Pixel phone will cost more than the Pixel 10. Barkat said Google had "shielded consumers from supply fluctuations for as long as possible" but that "economics have fundamentally shifted." Apple, Nintendo, Microsoft, and Roku have all raised prices recently for similar reasons.
Hugging Face CEO demands 'radical transparency' after OpenAI hack claim
Hugging Face CEO Clément Delangue has called for "radical transparency" from AI companies following a security incident affecting OpenAI that he described, in his own words, as "the first autonomous agent cyberattack." Delangue called the event "unprecedented" and said it "deserves an unprecedented response." Those characterizations come from Delangue's public statement, not an official OpenAI account of the incident.
Token efficiency: why it matters more than raw AI capability
New AI models increasingly compete less on dramatic capability leaps and more on doing the same job with fewer "tokens" at lower cost. One company's latest flagship model has become a case study in that shift, matching rival output at a lower price rather than topping every benchmark. This explainer breaks down what token efficiency means and why the industry's focus is moving there.
Chinese AI models: why releases like Kimi keep rattling Wall Street
Moonshot AI's release of its Kimi model triggered a familiar bout of anxiety in Silicon Valley and on Wall Street, echoing earlier Chinese AI launches that undercut US pricing. The pattern says less about any single model and more about how much AI valuations still depend on durable pricing power. This explainer breaks down why cheaper, competitive Chinese models keep unsettling US markets.
AlphaFold AI Helps Researchers Redesign Gene-Editing Proteins for Safety
Researchers are using Google DeepMind's AlphaFold, an AI system that predicts protein 3D structures, to identify which parts of gene-editing proteins cause off-target errors. The work aims to redesign CRISPR-associated proteins so they cut DNA more precisely, reducing mistakes that could affect unintended genes. The research is still at the protein-engineering stage, not a clinical treatment.