European Researchers Blocked From Auditing Social Media Algorithms Despite Stricter Digital Services Act Rules Academic institutions across Europe face continuous obstacles and legal hurdles as social media giants restrict API data access needed to study digital election interference. An unexpected and highly volatile political wave recently swept through Romanian politics. Hardline views targeting immigration alongside antisemitic rhetoric gained immense traction online, with posts from Calin Georgescu reaching over 120 million views before voters headed to the polls. Previously polling in single digits, Georgescu sent shockwaves through the political landscape by securing 23 percent of the vote in the opening round of the presidential race. This sudden rise left observers scrambling to understand how a previously fringe figure could dominate national discourse so rapidly. The Stumbling Block of Platform Access In the Netherlands, Adriana Iamnitchi, who chairs computational social sciences at Maastricht University, wanted to understand the mechanics behind this viral surge. Her team specializes in analyzing coordinated online disinformation campaigns. Specifically, she aimed to uncover how pro-Georgescu messaging was being monetized on TikTok through undisclosed influencer partnerships and live political broadcasts. On October 28, 2025, exactly a month prior to Georgescu's first-round victory, her team formally requested access to TikTok's Application Programming Interface (API) using the provisions of the European Union's Digital Services Act (DSA). Their request met a dead end. TikTok turned down the application, stating that the university team had failed to sufficiently prove their standing as established researchers, clarify potential commercial interests, or meet the platform's stringent data security standards, as Iamnitchi later shared in a detailed blog post. This refusal effectively blocked independent oversight during a critical electoral window. Belated Moderation and Disrupted Elections Two months after the rejection, TikTok identified 116,000 potentially compromised accounts and dismantled a network of more than 27,000 fake accounts. This coordinated network, operated by a third-party fake engagement provider, was actively boosting Georgescu and his political party, the Alliance for the Union of Romanians (AUR). TikTok admitted it was unable to trace the origin or operators of the campaign. The first round of voting was subsequently annulled, leading to a May 2025 runoff where independent candidate and former Bucharest mayor Nicușor Dan defeated George Simion, who took charge of AUR after Georgescu was disqualified. Georgescu denounced the court's annulment of the 2024 vote on his YouTube channel, calling it a formalized coup d'état by Romania's Constitutional Court. Subsequent declassified intelligence reports from Romania suggested Georgescu benefited from preferential treatment and artificial amplification on TikTok, with allegations pointing toward coordinated Russian involvement. Iamnitchi noted that had her team obtained the requested data, they might have traced the entities funding and organizing the campaign. She expressed frustration that when crucial societal research depends on privately held platform data, investigating these digital phenomena becomes almost impossible. The Decline of Public Auditing Tools Researchers across the continent face similar patterns of resistance and obfuscation. Despite legal entitlements granted by EU legislation, academics often have to resort to litigation to get essential data. While regulatory frameworks were recently expanded to improve transparency, researchers state that platform compliance and implementation remain the primary obstacles. Over the last few years, major platforms have retired widely used auditing tools like Meta's CrowdTangle, replacing them with restrictive Content Libraries. Meanwhile, X placed its API behind expensive paywalls, costing academic projects hundreds of dollars a month, according to Duncan Allen, a research officer at Germany's Democracy Reporting International (DRI). This lack of access prevents researchers from auditing recommendation algorithms or assessing how platforms handle sensitive content reports. Additionally, tight daily limits on data downloads, such as those imposed by TikTok, make large-scale, automated analysis mathematically unfeasible for academic teams. Technical Barriers and Data Quality Issues The DSA was designed to grant verified researchers at reputable institutions access to raw platform data to study systemic threats, including hate speech, election interference, and civil rights violations. However, the physical infrastructure requirements demanded by platforms present major operational challenges. Iamnitchi pointed out that many platforms require data to be stored on highly secure, air-gapped systems completely isolated from the internet, which standard university labs simply do not possess. Even when research teams obtain approval, the utility of the provided data is often poor. L. K. Seiling, who coordinates the German-based DSA40 Collaboratory, noted that API datasets are frequently inconsistent and difficult to replicate, preventing the scientific peer-review process. Out of 46 applications tracked by the group, 20 were approved and 14 rejected. The rejection rates varied greatly between platforms; TikTok accepted 11 of 13 requests, whereas X turned down 11 of 23. Seiling observed that the current data access framework seems designed to actively discourage academic scrutiny. In response, a TikTok representative stated the company has supported over 1,500 research teams globally, including 130 approvals within the EU during the latter half of last year. The company maintained that its daily API limit of 1,000 requests, which yields up to 100,000 video records or 2 million follower data points, is fully compliant with the DSA. Similarly, a Meta spokesperson defended the transition from CrowdTangle, asserting that the new Meta Content Library offers the most comprehensive data coverage across Facebook, Instagram, WhatsApp, and Threads, while maintaining high privacy standards. Legal Recourse and Regulatory Penalties With few automated alternatives, researchers often rely on manual data scraping, which exports basic web data into spreadsheets. However, Allen emphasized that scraping cannot map complex, coordinated disinformation networks because it fails to capture complete follower structures. Consequently, researchers have begun taking platforms to court. DRI and the Society for Civil Rights sued X in February 2025 after their April 2024 request to study German political discourse was rejected. Although a Berlin court ruled in their favor, they faced a similar roadblock during Hungarian elections, where a jurisdictional dispute almost derailed the case before being won on appeal. Frustrated by these persistent barriers, the European Commission took action in December 2025 by fining X €120 million (approximately $137 million) for creating artificial hurdles that hindered public interest research. X appealed the ruling on February 20, 2026, claiming the investigation violated procedural rights. However, the commission recently accepted a compliance plan from X to streamline researcher vetting, remove scraping bans, and offer free data access within six months. While researchers welcome these steps, they remain cautious about how effectively these platform promises will translate into actual scientific access. What this means for you • On the General Public: Platform data restrictions make it harder for experts to spot election manipulation and fake news, meaning the information you see online could be silently biased. • Accountability and Safety: Without academic audits, tech giants face less pressure to secure their algorithms, reducing overall user privacy and safety online. Questions & Answers 1. Why were European researchers requesting data from TikTok and Meta? Researchers were seeking data to study how coordinated networks of fake accounts and hidden influencer marketing were used to spread political disinformation during the Romanian elections. 2. What is the Digital Services Act (DSA)? The DSA is a European Union regulation that legally requires large online platforms to share API data with vetted academic researchers to study systemic risks like hate speech and election interference. 3. What technical difficulties do researchers face when trying to access platform data? Platforms often demand that universities store retrieved data on highly secure, air-gapped computers completely disconnected from the internet, which most academic departments cannot afford or maintain. 4. Why did the European Commission penalize X? In December 2025, the European Commission fined X €120 million for creating unnecessary barriers that undermined researchers' legal rights to audit the platform under EU law. https://trendkia.com/en/technology/eu-ke-sakht-niyamon-ke-bawajood-social-media-par-dushprachar-ki-janch-karne-se-roke-ja-rahe-researchers-10345 TrendKia — Har trend, sabse pehle.