The aim of the project was to develop a trustworthy AI component. This is intended to support both whistleblowers and employees of a whistleblowing system through partial automation.
Employees or other parties involved often do not report violations such as non-compliance with internal guidelines, bullying or sexual harassment. The reason often lies in the fear of negative consequences. These include discrimination, sanctions or even loss of job.
A trustworthy reporting channel is therefore essential. It must enable high-quality reports that support the rapid clarification of incidents and at the same time protect the whistleblowers.
However, evaluating reports via conventional channels often means a lot of manual work for case handlers. The quality of the texts is rarely sufficient, and whistleblowers are usually only available for queries to a limited extent. In addition, there are reservations about the high costs of analysis and the risk of misuse, for example through false reports. These factors often deter companies from introducing a whistleblowing system.
EU Directive 2019/1937 on the protection of whistleblowers must be transposed into national law by December 2021. Compliance 2b GmbH, a start-up founded in 2020, has taken on this task.
With the Act4Whistleblowing web solution, it offers a secure, anonymous exchange of information between whistleblowers and companies. At the same time, the platform enables simple and innovative processing of reports by employees.
In a further expansion stage, RISC Software GmbH is developing a trustworthy AI component together with Compliance 2b GmbH. It supports whistleblowers in reporting and makes processing easier for case handlers. This allows reports to be processed more efficiently and with fewer errors.
The system automatically derives information from clues and compares it with existing structured data. This plausibility check opens up new possibilities:
The increased quality of the reports improves the chances of a successful investigation. At the same time, the risk of false reports or falsifications is reduced.
The research project was funded by Austria Wirtschaftsservice (“aws”) as one of 12 out of 68 submissions in the second round of the call “Digitalization – Trustworthy Artificial Intelligence”
Data Scientist
