Marie-Therese Sekwenz

Marie-Therese Sekwenz, Shreyan Biswas, Rita Hermann-Gsenger et al.

Illegal content reporting mechanisms are a key technical and organizational measure through which online platforms address illegal content under the European Union Digital Services Act (DSA). Article 16 requires user notices to be sufficiently substantiated and submitted in good faith, placing users in the difficult position of interpreting legal and procedural language and translating ambiguous content into legally meaningful categories and reasons. We investigate how large language model (LLM)-based assistants can support this reporting process. In a controlled user study (N = 450) using an interface modeled on a major platform reporting workflow, we compare three conditions: unaided reporting, a conventional explainable AI assistant (XAI) that suggests a single legal category with a rationale, and an evaluative AI assistant (EvalAI) that presents balanced pro and con arguments across candidate legal provisions. We further examine these assistance forms under systematically varied AI error regimes. Our results show that EvalAI improves provision-level accuracy under AI error and reduces misclassification distance relative to conventional XAI, particularly for near-miss and overbreadth errors. When AI output is correct, conventional XAI enables faster decisions, but neither AI assistance form reliably improves the quality of users' substantiated explanations relative to unaided reporting. We discuss design implications for compliance-oriented reporting interfaces, highlighting trade-offs between accuracy, deliberation, explanation quality, and vulnerability to misleading AI output.

Marie-Therese Sekwenz, Kyle Beadle, Simon Parkin

The European Union's Digital Services Act (DSA) introduced regulatory mechanisms which serve as a way to manage harmful content online. The recognition of Trusted Flaggers (TFs) is one such mechanism which accredits entities with experience, platform independence, and skill in identifying and reporting illegal content. With the DSA's TF role being roughly one year old, we interviewed representatives of seven such TF organizations to learn about their experiences of becoming a TF and how it impacts their interactions with online platforms and with individual users. We additionally ran a workshop involving TF representatives, primarily as it was requested by TFs themselves, who collectively wanted to share experiences of their new role and learn from each other rather than be isolated. Notably, we found that accreditation as a TF can be cumbersome, that resources for TFs remain the same despite an increasing workload, and that platforms priorities often diverge from TFs. We conclude with recommendations for future research into understanding user representation within the DSA and the need for standardization measures tailored to the needs and resource constraints of TFs.