Social Media Giants Accused of Failing to Curb Disinformation During 2025 Irish Presidential Election
A new report has revealed a significant failure by major social media platforms to act on flagged disinformation during the 2025 Irish presidential election, raising serious concerns about the integrity of online political discourse. The European Digital Media Observatory (EDMO) Ireland, which monitored the electoral period, documented a substantial number of policy violations that were largely ignored by the tech companies responsible for hosting the content. The findings have sparked a debate about the effectiveness of self-regulation and the need for more robust enforcement of digital rules, with the report explicitly criticizing the platforms for their inconsistent and opaque decision-making processes.
According to the report, EDMO Ireland identified a total of 177 instances of posts or activity that appeared to breach the social media companies’ own community guidelines and terms of service over the course of the election campaign. The breakdown of these violations reveals the multifaceted nature of the problem, with election-related disinformation being the most prevalent category. The report counted 85 specific cases of disinformation, alongside 26 instances of AI-generated election content. Other significant categories included 49 cases of illegal election-day content, 13 cases of harassment and abuse, and two separate incidents of incitement to violence. The final category, issues relating to political advertising, accounted for two of the flagged cases, highlighting the breadth of the challenges faced in regulating online political activity.
The most concerning aspect of the report is the low rate of action taken by the platforms in response to these clear violations. Of the 85 counts of electoral disinformation that were flagged to the companies, they only acted on a mere 12 of them. This paltry response rate demonstrates a systemic failure to uphold their own stated policies and protect the democratic process from manipulation. The report suggests that this inaction is not due to a lack of effort on the part of watchdogs, but rather a deliberate or negligent choice by the platforms, which have the technical capability to remove or moderate content but failed to do so in the vast majority of cases. This has led to serious questions about the commitment of these corporations to ensuring the integrity of elections in the digital age.
The report highlighted several specific examples where platform inaction was particularly glaring. One of the most prominent cases involved deepfakes of then-presidential nominee Catherine Connolly that were circulating online. The fabricated videos, which falsely announced her withdrawal from the election, gained widespread attention in the week leading up to the vote. Despite being a clear attempt to deceive voters and interfere with the election, the platforms reportedly failed to act swiftly or decisively on these dangerous AI-generated contents. Other incidents flagged by EDMO Ireland included photos of marked ballots posted on election day, which is illegal under Irish law. In these cases, the three faces on the ballot papers were often defaced, yet the content remained live. The report also pointed to a video of former Taoiseach Leo Varadkar being harassed in Dublin, which was flagged but not adequately addressed.
In its detailed findings, EDMO Ireland singled out specific platforms for their poor performance, with Facebook and TikTok identified as the “worst offenders.” Facebook had 54 cases flagged to it but only acted on 44% of them, while TikTok had 50 cases but only actioned 54%. The report was particularly critical of Facebook, noting that its failure to moderate content was surprising given that its parent company, Meta, had a formal partnership with the Irish digital news site, The Journal, for fact-checking purposes. The report highlighted a specific case where a false story, based on a fabricated quote attributed to a High Court judge, was posted on Facebook. The quote called into question the constitutional validity of the nomination process. Despite six items containing this claim being reported to Meta, along with a fact-check from The Journal, the platform repeatedly declined to take action, stating without further elaboration that the content did not violate its terms of service.
While the performance of Facebook and TikTok was heavily criticized, other platforms fared better in relative terms, though their overall contribution to the problem was still significant. LinkedIn, for instance, had seven cases flagged to it, and notably did not act on a single one of them, meaning it had a 0% action rate. This suggests that the problem is not confined to the largest social networks but is a systemic issue across the industry. On the other hand, TikTok said that 97.5% of violating content was removed before it was reported, indicating that its algorithmic detection systems were working to some degree. However, the platform’s response to items flagged by external monitors was still seen as insufficient. Similarly, YouTube stated that it “reviewed all reports received and removed those that violated our policies,” but the EDMO Ireland report suggests that the overall picture remains one of widespread failure to protect voters from misinformation and manipulation.
The report’s release has brought the effectiveness of the EU’s regulatory framework into sharp focus. The EU’s Digital Services Act (DSA) recommends that platforms take proactive measures to direct users to official election information, demonetise disinformation, and reduce its visibility. It also calls on them to tackle fake accounts and help users evaluate information. EDMO Ireland was involved in a ‘Rapid Response System’ during the election, a collaborative effort with participating platforms designed for self-regulation. However, the report notes that X (formerly Twitter) had pulled out of this commitment, and that even the platforms that remained in the system failed to provide expeditious responses as agreed. The core finding of the report is that platforms were “repeatedly unable to explain inconsistencies in the application of their own rules,” and that “signposting to authoritative sources, corrective information, and other proactive information integrity efforts were poor across all platforms.” This damning conclusion suggests that the current approach to self-regulation is failing, and that stronger, more mandatory enforcement mechanisms may be required to hold these powerful tech companies accountable for their role in safeguarding democratic elections.