Is the Web Closing? Perspectives from OxyCon 2026

Litigation is the third signal

Three perspectives on the same problem

Krishna Sood, who leads product legal at Black Forest Labs (the Freiburg-based company behind the FLUX models), spoke as one of the few people building frontier AI in Europe. 

Europe’s research-access gap

This part of the discussion cuts to the heart of why public-interest research is struggling. EU law, the Digital Services Act, and UK law, the Online Safety Act, both require major platforms to give vetted researchers access to data for public-interest research. Carl mentioned, "To answer plainly: no, it is not happening effectively." Researchers face endless onboarding delays, broken APIs, and unresponsive channels, and in the UK the implementation guidelines have stalled.

The knock-on effect is that researchers can't even assess whether platform-provided data would be trustworthy, because they can't get it in the first place. That keeps independent scraping essential to even begin to answer urgent questions, like whether algorithmic feeds amplify extremist content.

Paul gave a concrete example: an ITV journalist, a former student of his, partnered with Oxylabs Project 4β to gather public data during the UK civil unrest following the tragic deaths of children, tracing how misinformation spread. He also drew a parallel to the Open Data movement 15 years ago, when governments released the datasets they were comfortable with and kept the sensitive ones back.

The problem with licensing everything

One proposed fix is moving entirely to commercial licensing, and the panel pushed back. Krishna argued that licensing makes sense for private, proprietary assets, but public web data should stay open for analysis. 

This is bigger than business

Carl closed with the heaviest point. About a decade ago, defense organizations started treating information as a theater of war alongside land, sea, air, and space. Hostile state actors, including Russian and Chinese hybrid operations, see the information domain as a key arena for weakening democracies. That fight used to play out on social media; now AI systems are the next target, probed and manipulated at scale to shift public perception. 

Key takeaways

The takeaway from the panel was clear: we have to keep making the case for public data access, out loud and often. This isn't a fight you win once and for all. Pressure from courts, policy, and technical barriers keeps building, and we have to make the case for open access again and again, in rooms where these decisions actually get made.

The web may not be closing for everyone at once, but the people who use it to hold power to account, study real harms, and fight misinformation are feeling the doors swing shut. Those are some of the uses that matter most for the public good, and they need the commercial muscle to push back together.

Join 4β

That gap is exactly why Project 4β exists. We provide free public web data collection tools, technology, and legal advisory to researchers, journalists, academics, NGOs, and nonprofits working on public-interest missions, so the people doing this work aren't the ones locked out.

Much of it falls within the areas this panel covered. On information warfare, democratic resilience, and investigative journalism, we already work with organizations like Bellingcat, Global Witness, Debunk.org, Doublethink Lab, and the Civic Resilience Initiative. On the academic side, we support researchers at universities including Stanford University, Carnegie Mellon University, Northwestern University, and Warwick University.

If your work depends on getting to public web data, whether you're tracing how false narratives spread, auditing algorithmic harms, or building research that defends open democracies, we'd like to help. Academics, NGOs, NPOs, and other mission-driven organizations can join the initiative and get free access to tools that match the scale of what they're trying to do.

Tell us about your data needs through our form or via form or at 4beta@oxylabs.io.