Fraud-Detection – automatisierte Erkennung gefälschter Dokumente

Österreichische Post AG is repeatedly confronted with manipulated documents. In addition to classic, manually produced forgeries, there is an increasing number of documents that have been manipulated using image and text editing as well as AI – these are often altered PDF, image, or text files. The rapid development of technological possibilities means that these forgeries are difficult or impossible to detect with the naked eye. At the same time, the current manual checking of documents involves considerable time and financial expenditure.
Another major challenge is the sheer volume of documents that Österreichische Post AG has to check. Every year, several million documents have to be checked for authenticity, which is only possible through complete automation using AI-supported processing. Intelligent Document Processing will enable the extracted and classified documents to be reliably identified as forgeries. Intelligent document processing is intended to extract and classify the documents received in order to reliably identify forgeries. Österreichische Post AG is therefore looking for a scalable, AI-supported solution for intelligent document processing in the area of fraud detection that can be integrated into existing workflows and systems.

The solution should be able to identify forgeries with a high degree of reliability and accuracy.

Österreichische Post AG is seeking an automated, AI-supported fraud detection solution that reliably detects document forgeries, can be integrated into existing workflows, and can handle large volumes of documents. The goal is to quickly and, above all, reliably identify forged documents using image and text recognition, regardless of whether they are analog or digital documents. The verification process focuses on documents from the human resources department, such as pay slips and residence registration forms, as well as accounting documents, such as bank statements, or invoices that have been manipulated manually or using software. The documents are to be checked for irregularities such as structural breaks, metadata inconsistencies, layout anomalies, font deviations, retouched areas, or inconsistent data. It is important that AI-generated content and deepfakes can also be detected. The solution sought should support existing processes, but not make final decisions (no straight-through processing).

High scalability is particularly important for the system we are looking for in order to efficiently handle the annually growing document volume of several million pages. The operating and support model should ensure that regular updates can be used to respond to new fraud attempts. Furthermore, the solution sought must be fully integrable into existing workflows, which is why modular systems or low-code/no-code solutions, for example, are also of interest. In addition, the software sought should be open to various operating models and APIs in order to integrate it as well as possible into different workflows.

Another key aspect concerns the possibility of KPI-based evaluations. To this end, the solution should have a clearly structured dashboard that shows in an understandable way how many forgeries have been identified in the various departments and what proportion of the documents checked can be classified as forged. This creates both transparency and a sound basis for decision-making. Solutions with a German user interface are also desired. English elements can be offered as an option.
There is flexibility with regard to the technical design: The solution sought can be offered as both a cloud variant and an on-premise solution, provided that it can be guaranteed that even large volumes of data can be processed reliably within a short time. Integratable systems as well as complete SaaS solutions can also be submitted.

As Österreichische Post AG is a state-owned company, the solution submitted must comply with the GDPR and meet the latest security standards to ensure data security. In addition, the systems must withstand audits and certifications (e.g., ISO 27001, 9001, and other data protection and compliance-related certifications). This requires that data is processed exclusively in data centers within Europe and that no data is transferred to third countries. This also applies to data transfer by subcontractors or distributed cloud structures. In addition, all technologies used must be EU-compliant, particularly with regard to data protection and cloud hosting. If cloud systems are used, a hosting solution within the EU is mandatory.

In addition to high data security, the transparency of the system and the price transparency of the submission are also of great importance.
We are looking for a fully functional product; pure concepts are not desired. If you are invited to the innovation dialogue, you will be provided with a set of documents to test the software in practice.

Important: Additional technical requirements to be taken into account are available in the download area below.

Not objectives:
• Submission of a pure prototype
• Solutions that cannot be integrated into existing workflows

Take advantage of this opportunity to present your innovative approaches. Österreichische Post AG is looking for companies or cooperation partners that offer scalable, workflow-integrated, and AI-supported solutions for intelligent document processing in the area of fraud detection. Companies should submit their solutions using the online form by November 9, 2025.

Description
Focus primarily on the technical solution you would use to achieve the desired result. Describe the overall approach of your solution, how it works, and how ongoing support and training costs are taken into account in addition to technical implementation.
The focus of the description should match the evaluation criteria and the requirements described.
Convince the reader with comprehensible use cases, key figures, and, if possible, even references.

Summary of Added Value:
Highlight your unique selling proposition and key advantages! What makes your solution particularly innovative? What sets you apart? Use the technical information provided and indicate whether you can meet these requirements.

Confidential information (for the jury only):
Give us an idea of what your proposal means in terms of costs and effort. This will help the sponsors to design the procurement project within the right framework at an early stage. Please describe this information with reference to the project.
In this text field, you can enter additional information that you would like to share exclusively with the project managers. Alternatively, you can send a PDF by email to the challenge moderators.

In addition, a cover image is required. This is your visual flagship in the overview of submissions for this challenge.

Optionally, you may also attach files or graphics that complement, but not duplicate or replace, the texts in the form fields!

IMPORTANT NOTE:
Submissions as a consortium are possible. Please be brief (guideline: maximum 10-12 presentation slides or 4-5 DIN A4 pages). This challenge is part of a market exploration. It is therefore not necessary to prepare specially developed drafts or feasibility studies for this occasion.