CoP Monitor: Baseline Reports: Assessment of VLOP and VLOSE Signatory Reports for the Strengthened Code of Practice on Disinformation

After a revision of the 2018 Code, the “Strengthened Code of Practice on Disinformation” was adopted in 2022 by 34 (now 44) Signatories including major platforms. The updated Code introduces stronger monitoring requirements and reporting through quantitative Service Level Indicators (SLIs) and Qualitative Reporting Elements (QREs), increased support for fact-checking, the establishment of a Transparency Centre for Signatory reporting and commits Signatories to collaborate and continuously update the Code through a Permanent Task-force.

After a revision of the 2018 Code, the “Strengthened Code of Practice on Disinformation” was adopted in 2022 by 34 (now 44) Signatories including major platforms. The updated Code introduces stronger monitoring requirements and reporting through quantitative Service Level Indicators (SLIs) and Qualitative Reporting Elements (QREs), increased support for fact-checking, the establishment of a Transparency Centre for Signatory reporting and commits Signatories to collaborate and continuously update the Code through a Permanent Task-force. This analysis examines the baseline reports published by the largest Signatories of the Strengthened Code in February 2023: Google, Meta, Microsoft, TikTok and Twitter/X. So far, no comprehensive, in-depth analysis of these reports has been published. Through a robust assessment scheme and three-stage coding process, we evaluated both qualitative and quantitative information provided by Signatories and assigned an overall score of 1 (poor), 2 (adequate) or 3 (good) for each Measure.

Our findings show that the overall quality of reporting is less than adequate. More than half of the applicable Measures exhibited incomplete or missing qualitative information. Notably, quantitative data was absent in 64% of relevant instances, and when provided, the methodology was sometimes questionable with data often imprecise or missing on a Member State-level. Furthermore, Signatories frequently included information unrelated to the specific Measure in question. In the "Empowering Research" section, Signatories' performance was particularly poor, indicating a need for substantial improvement in this area.

Our analysis highlights a discrepancy between a formal, report-centred approach to monitoring and a more substantive, investigative approach. The former focuses on verifying the fulfilment of reporting obligations, while the latter delves into the veracity and comprehensiveness of the information provided by Signatories. The substantive approach demands greater resources, expertise, and data access but is essential for meaningful monitoring.