The prevalence of internet fraud in Austria has been steadily increasing. According to the Cybercrime Report from the Federal Minsitry of Interior (BMI), 34,069 cases of internet fraud were reported in 2023, an increase of 23.3% compared to the previous year. Social media platforms are increasingly becoming a gateway for cybercriminals, leading to increased financial losses - particularly regarding fraudulent investment advertisements. Legal remedies have, as of now, not been effectively implemented.

Making use of the new regulatory instruments of the Digital Services Act (DSA), AdWatch contributes to improving security in Austria by effectively combating fraudulent advertising through the following measures:

(1) Establishing a knowledge base pertaining to the modi operandi and extent of fraudulent advertising as well as on criminal networks and vulnerabilities. The knowledge base encompasses the evaluation of existing suspected cases, the (cluster-) analyses of further cases, analyses of scope, targeting metrics, criminal strategies and the machine learning-based identification of fraud indicators, which supports the early detection of newly emerging fraudulent advertising. The results go beyond previous individual case analyses on the topic and will be published in a study report.

(2) Partial automation of relevant workflows for the detection, analysis, documentation and reporting of suspicious and fraudulent advertisements. In close cooperation with the Federal Ministry of Interior (BMI), workflows are defined that cover the process from detecting and analyzing suspicious cases to their documentation and reporting. The development of a robust system that partially automates these defined workflows is facilitated by the use of the API interfaces of the advertising libraries of very large platforms and search engines (VLOPs, VLOSEs), which have become mandatory under the DSA. The system is designed to support court-proof documentation of evidence in cases of fraud and the rapid removal of fraudulent advertisements.

(3) Development of recommendations to improve DSA compliance. Drawing upon the findings of legal analyses and empirical evaluations, the novel DSA approaches are evaluated in terms of their effectiveness. In turn, policy recommendations are derived. Analyses will be carried out on transparency obligations with regard to advertising libraries, reporting processes, systemic risks, and the platforms' mitigation strategies.

(4) Supporting authorities in law-enforcement. Firstly, the results support the authorities in law enforcement, secondly, they are aimed at the supervisory authorities of the DSA (European Commission, KommAustria). Other stakeholders are supported in their advisory, prevention, and policy work. Dissemination and awareness work will be conducted via the ÖIAT channels (e.g. Watchlist Internet) to protect consumers in a target group-oriented manner from fraudulent advertising.

The project is led by ÖIAT, which has successfully implemented numerous research projects on cybercrime and automated fraud detection. AIT's expertise in AI and machine learning and TU Vienna's experience in the development of crawler systems complement these competencies. The synergy between the partners is set to establish a foundational framework for the systematic identification and mitigation of fraudulent advertising practices, thereby contributing to enhancing trust in digital infrastructure and the reduction of economic losses.