Netherlands-based FRISS has partnered with Microsoft to improve its artificial intelligence (AI) fraud model explanations leveraging Microsoft Azure OpenAI service.
A provider of trust automation solutions for property and casualty (P&C) insurance carriers, FRISS intends to implement Azure OpenAI Service for multiple use cases within its offerings.
Some of the use cases include a commercial underwriting chatbot or a claim summariser, a network explainer, and using a fraud bot.
The solutions offered by FRISS will be available in Microsoft Commercial Marketplace in the form of a transactable offer in Azure Marketplace and Microsoft Appsource storefronts.
Currently, the Dutch company is working to create a ‘smart’ explanation of its AI models’ predictions, while ensuring a responsible AI approach.
FRISS said its platform aligns with the seven principles outlined in the “ethics guidelines for trustworthy AI” developed by the European Commission’s High-Level Expert Group.
How well do you really know your competitors?
Access the most comprehensive Company Profiles on the market, powered by GlobalData. Save hours of research. Gain competitive edge.
Your download email will arrive shortly
Not ready to buy yet? Download a free sample
We are confident about the unique quality of our Company Profiles. However, we want you to make the most beneficial decision for your business, so we offer a free sample that you can download by submitting the below formBy GlobalData
With Azure OpenAI Service, FRISS can create more engaging and easier-to-comprehend cases on potential fraud instances using storytelling to give explanations.
FRISS co-founder and chief strategy officer Christian van Leeuwen said: “Every time our AI model makes a prediction, we automatically create sentences, or bullet points, which explains why the model made such a prediction.”
This approach is expected to help insurance professionals and fraud investigators to understand potential fraud incidents in a more seamless manner, FRISS noted.
In addition, FRISS’ decision to use ChatGPT for explanations only will help in minimising potential bias or ethical concerns, as it does not influence the fraud assessment.
FRISS Data Science product owner Gian Luigi Chiesa said: “After conducting an initial survey, the response to the ‘smart story’ format was very positive.
“We are now working on implementing this approach for a group of pilot customers, ensuring its seamless integration and effectiveness.”
Leeuwen added: “As we move forward, we envision even more possibilities, with plans to release use cases for our customers based on generative AI, such as large language models, into our platform in the first half of 2024.”