The EU Digital Services Act Is the Best Experiment We Have, But Transparency Is Not Accountability

The EU Digital Services Act is the most important platform governance experiment in the world right now. That does not mean it is perfect. It means it is serious.

The DSA moves platform accountability beyond voluntary transparency and into structured obligations. It requires due diligence from online intermediaries, imposes special obligations on very large online platforms and search engines, and treats systemic risk as a regulatory concern rather than a public-relations inconvenience. Very large online platforms and search engines are designated when they reach more than 45 million monthly users in the EU. The law requires risk assessment and mitigation around issues such as illegal content, fundamental rights, civic discourse, electoral processes, public security, and other systemic risks.

This is a major shift. It recognises that platforms are not just hosting speech. They are structuring attention, visibility, monetisation, amplification, and access to public life.

The DSA also matters because it pressures platforms to provide data access for scrutiny. The European Commission has taken action around platform transparency obligations, including preliminary findings against TikTok and Meta over issues such as researcher access, notice-and-action mechanisms, and user appeal systems. These cases show why the DSA is not merely symbolic. It gives regulators a way to test whether platforms are making accountability possible in practice.

But the DSA’s greatest strength also reveals its greatest weakness: it is heavily dependent on transparency.

Transparency is necessary, but it is not enough. A transparency report can become a compliance costume. A risk assessment can become a polished PDF. An ad library can exist while remaining hard to use. A researcher-access system can be legally available but practically slow, narrow, or exclusionary. A complaint mechanism can technically exist while being designed in a way that discourages use.

The DSA will be judged not by whether platforms publish more documents, but by whether those documents change power.

There are four accountability tests the DSA must pass.

First, the usability test. Can researchers, journalists, civil society groups, and regulators actually use the information provided? Transparency that requires elite technical capacity, insider legal knowledge, or months of access negotiation will reproduce the same power imbalance it claims to solve.

Second, the comparability test. Can risks be compared across platforms, countries, and time? If each company defines risk categories differently, transparency becomes a pile of bespoke corporate narratives. Regulators should push toward standardised reporting formats, common metrics, and clear explanations of methodology.

Third, the remedy test. Does transparency help affected users? If someone’s content is removed, their account is restricted, their community is targeted, or their election environment is manipulated, can they understand what happened and challenge it? The DSA must connect systemic risk governance with user-level due process.

Fourth, the Global South spillover test. Does the DSA improve platform accountability outside Europe? Platforms may build better systems for the EU while leaving weaker protections elsewhere. This would create a regulatory caste system: European users get rights-grade accountability, while users in fragile democracies get diluted safety theatre.

The last point matters especially because the harms the DSA addresses are not European-only harms. Election manipulation, harassment, hate mobilisation, discriminatory enforcement, opaque advertising, and platform data scarcity are global problems. If DSA compliance infrastructure becomes available only where regulators are powerful enough to demand it, the law may deepen global inequality in platform accountability.

There is also a political risk. Critics of the DSA sometimes frame it as censorship. Some of that criticism is opportunistic, especially when coming from actors who prefer platforms to remain unaccountable. But the concern cannot be dismissed entirely. Any platform regulation that touches content governance must protect freedom of expression, procedural fairness, independent oversight, and proportionality.

This is why the DSA’s legitimacy depends on due process as much as enforcement. Strong platform regulation should not mean states quietly pressuring companies to remove speech. It should mean platforms and regulators being forced to explain their decisions, expose their assumptions, protect rights, and accept scrutiny.

The best version of the DSA would become a model for evidence-based platform governance. It would make systemic risk assessments meaningful. It would make data access real. It would force platforms to explain how recommender systems, advertising systems, moderation systems, and reporting systems affect public life. It would give researchers enough access to test corporate claims. It would create pressure for other regions to demand equivalent standards.

The worst version would create paperwork accountability: more forms, more reports, more controlled access, and more regulatory theatre, while ordinary users still face opaque decisions and civil society still struggles to document harm.

The difference between those futures is enforcement design.

Transparency asks platforms to show something. Accountability asks what happens when what they show is inadequate.

That is the DSA’s real test.