The Digital Services Act’s Broken Promise
The European Union’s Digital Services Act (DSA), enacted to curb online harms and ensure transparency, is facing its first major stress test. Researchers across Europe are sounding alarms: TikTok, Meta, and X (formerly Twitter) are systematically obstructing access to platform data, despite legal obligations to provide it. The stakes couldn’t be higher. In the 2024 Romanian presidential election, a Russian-coordinated disinformation campaign allegedly boosted candidate Calin Georgescu to 23% of the vote—largely through TikTok. Yet when researchers like Adriana Iamnitchi of Maastricht University sought data to study the campaign’s impact, they hit a wall.
The DSA was designed to prevent exactly this scenario. Under Article 40, platforms must grant researchers access to “public data” to study “systemic risks,” including disinformation. But as L. K. Seiling of the DSA40 Collaboratory puts it, “Data access as it’s set up right now tries to disincentivize researchers.” The result? A regulatory framework that looks good on paper but fails in practice.
How Big Tech Is Gaming the System
TikTok: The Quota Trap
TikTok’s API is a case study in bureaucratic obstruction. While the platform touts its 1,000 daily API requests and capacity to pull 100,000 video/comment records, researchers report that these limits are arbitrarily low for large-scale studies. Iamnitchi’s team, investigating the Romanian election, found that 120 million views of Georgescu’s posts couldn’t be fully analyzed due to these caps. TikTok claims to have approved 130 EU applications in the second half of 2025, but the DSA40 tracker shows only 11 out of 13 applications were approved—suggesting a 84.6% approval rate, but with crippling restrictions.
Meta: The CrowdTangle Replacement Shell Game
Meta’s CrowdTangle, a once-powerful tool for tracking viral content, was shut down and replaced with the Meta Content Library and API. While Meta frames this as an upgrade, researchers argue it’s a downgrade in functionality. The new system requires nonprofit status or journalistic credentials, narrowing access. Worse, it lacks the granularity of CrowdTangle, making it harder to map disinformation networks. For a company that profits from engagement-driven algorithms, this opacity is convenient.
X: The Paywall Barrier
X’s approach is the most brazen. After paywalling its API in 2023, the platform now charges researchers “hundreds of dollars a month” for data access. Democracy Reporting International (DRI) sued X in February 2025 after its application was rejected in November 2024. The European Commission fined X €120 million ($137 million) in December 2025 for these barriers, but X appealed the decision in February 2026. The message is clear: X would rather pay fines than comply.
Why This Matters: Disinformation as a Weapon
The Romanian election is a microcosm of a global problem. A Russian-backed TikTok campaign allegedly amplified Georgescu’s candidacy, flooding the platform with 116,000 potentially compromised accounts and 27,000 fake profiles. When researchers tried to study the network, TikTok’s API quotas made it impossible. This isn’t just an academic issue—it’s a national security threat.
Disinformation campaigns don’t just sway elections; they erode trust in institutions. In 2024 alone, the EU identified dozens of coordinated inauthentic behavior (CIB) networks across social media. Yet without unfettered data access, researchers can’t:
- Map the spread of false narratives in real time.
- Identify bot networks before they influence public opinion.
- Hold platforms accountable for algorithmic amplification of harmful content.
As Iamnitchi warns, “When you need data to investigate that impact, and that data is privately held, it can become practically impossible to research this space.”
The Technical Breakdown: What Researchers Need vs. What They Get
| Requirement | TikTok’s Offering | Meta’s Offering | X’s Offering |
|---|---|---|---|
| Data Scope | Video, comments, follower records | Facebook, Instagram, WhatsApp, Threads | Public posts, limited engagement data |
| Access Method | API (1,000 requests/day) | Meta Content Library & API | Paywalled API |
| Approval Rate | 84.6% (11/13) | Not disclosed | 52.2% (12/23 rejected) |
| Cost | Free (but capped) | Free for nonprofits/journalists | “Hundreds of dollars/month” |
| Web Scraping Alternative | Blocked | Blocked | Blocked |
The Web Scraping Workaround (That Doesn’t Work)
Some researchers resort to web scraping, manually exporting data into spreadsheets. But this method is woefully inadequate:
- No follower lists: Critical for mapping coordinated networks.
- No engagement metrics: Impossible to track virality.
- Legal risks: Platforms like Meta and X actively sue scrapers under anti-bot laws.
The Industry Impact: A Chilling Effect on Research
The DSA was supposed to be a model for global tech regulation. Instead, it’s becoming a cautionary tale about corporate resistance to transparency. The consequences are already visible:
Delayed Responses to Disinformation
- Without real-time data, researchers can’t flag emerging threats until after they’ve gone viral.
- Example: The 2024 Romanian election disinformation was only fully understood months after the vote.
A Two-Tiered System
- Well-funded institutions (e.g., Stanford Internet Observatory) can afford X’s API fees.
- Smaller NGOs and academics are locked out, creating a knowledge gap.
Regulatory Capture
- Platforms lobby to define “systemic risk” narrowly, excluding many disinformation studies.
- The European Commission’s €120M fine on X was a rare win—but X’s appeal shows Big Tech’s willingness to fight compliance.
Global Ripple Effects
- The U.S. is watching closely. If the DSA fails, American regulators may abandon similar efforts.
- For more on how corporate lobbying shapes tech policy, see our analysis of Why Corporate AI Super PACs Spent $27M on a Local Election .
The Future Outlook: Can the DSA Be Fixed?
The DSA’s data access provisions are not inherently flawed—but their enforcement is. To fix this, the EU must:
1. Mandate Standardized APIs
- Platforms should be required to provide identical data fields (e.g., follower networks, engagement metrics).
- Example: TikTok’s 1,000-request daily cap should be scaled to research needs.
2. Create a Centralized Research Portal
- A single EU-run platform could aggregate data from all major social networks, eliminating individual paywalls and quotas.
- Precedent: The European Data Innovation Board could oversee this.
3. Strengthen Penalties for Non-Compliance
- The €120M fine on X was a start, but repeat offenders should face higher penalties (e.g., 5% of global revenue).
- Criminal liability for executives who willfully obstruct research should be considered.
4. Expand Legal Protections for Researchers
- Anti-scraping laws should include exemptions for academic research.
- Whistleblower protections for employees who expose platform interference.
5. Learn from Other Sectors
- Financial regulators require banks to provide real-time transaction data—why not social media?
- Healthcare data is shared under HIPAA with strict privacy controls. A similar framework could work for social media.
For a deeper dive into how tech platforms evade accountability, read our investigation into What Happens If China Hacks the US Water Supply? .
FAQ: What Researchers and Readers Need to Know
Q: Why can’t researchers just use public data?
A: Public data is often incomplete. For example:
- TikTok’s For You Page algorithm prioritizes content in ways that aren’t visible in public posts.
- X’s paywalled API hides engagement metrics (likes, retweets) that reveal bot activity.
Q: Isn’t this a privacy issue?
A: The DSA explicitly balances transparency with privacy. Researchers only access public data (e.g., posts, comments) and must anonymize user information. Platforms’ objections are disingenuous—they’re protecting business interests, not user privacy.
Q: What can researchers do now?
A:
- Apply for platform APIs (even if approval is unlikely).
- Collaborate with NGOs like Democracy Reporting International to pool resources.
- Pressure regulators to enforce DSA provisions more aggressively.
Q: How does this affect everyday users?
A: Disinformation spreads faster when researchers can’t study it. If platforms block data access, users are more likely to encounter:
- Fake news about elections, health, or crises.
- Coordinated harassment campaigns.
- Algorithmic amplification of extremist content.
For more on how algorithms shape public discourse, see our analysis of X’s Algorithm Update Prioritizing Replies .
Conclusion: The Battle for Transparency Is Just Beginning
The Digital Services Act was a landmark achievement—but its success hinges on enforcement. Right now, Big Tech is winning the war of attrition. Researchers are frustrated, underfunded, and outgunned by corporate legal teams. Yet the stakes—democracy, public health, and national security—couldn’t be higher.
The EU has two choices:
- Double down on enforcement, with stricter penalties and mandated data standards.
- Abandon the DSA’s transparency goals, letting platforms self-regulate (a proven failure).
The Romanian election disinformation scandal is a warning. If the EU doesn’t act, 2026’s elections—in the U.S., Germany, and beyond—could be even more vulnerable.
For now, the message to researchers is clear: Keep fighting. And to Big Tech: The world is watching.
Source: Original Article