Navigation

FIDA presents AI image forensics

At insureNXT in Cologne

SCROLL

At insureNXT in Cologne, our colleague and data scientist Gillian Kant gave a presentation on "Recognising synthetic multimodal data and its application in the insurance environment".

The increasing spread of disinformation through synthetically generated data poses a significant risk to society and the economy. Easy access to generative AI facilitates the spread of artificial images and texts, potentially misleading people. Therefore, the ability to recognise synthetic data and identify its characteristic features is of great importance.

In his presentation, he introduced our innovative application, FIDA Image Forensics, which distinguishes synthetic from real data. True to the motto: Trust is good, our image forensics is better. Gillian was accompanied by our division manager Karsten Völzke, who reported on his impressions of the event:

"Attending insureNXT was a first for me and I was impressed by the variety of topics presented. These ranged from the strategy of large insurers on the use of AI in companies, to training in companies and the presentation of specific use cases. I found Hamidreza Hosseini's masterclass on "Operational and practical fields of application of generative AI" particularly exciting.

Overall, I see this as confirmation that we are hitting the nerve of the times with both our AI training topics and our products. This was also made clear by the positive reactions to Gilian's pitch on the topic of "Synthetic data recognition and application in the insurance environment". Here, the results of the collaboration between the University of Göttingen and FIDA were presented, which led to our latest product, FIDA Image Forensics.

I am convinced that we will be represented at insureNXT again in 2025," concludes Karsten.

Cologne, May 14, 2024

Do you have any questions?

Finanz-DATA GmbH,
Helenenstraße 3, 99867 Gotha

Lisa Heß
Marketing Manager
Phone: +49 3621 4510-14
Mobile: +49 171 3306753
E-mail: marketing@fida.de