Munich Re has launched a new validation service for trustworthy artificial intelligence.
The firm says that the service, called CertAI, validates fully fledged AI systems or AI solutions already in productive use, analysing six different factors – fairness, autonomy and control, transparency, robustness, functional and cyber security, and data protection. It validates the trustworthiness of AI in new product solutions and demonstrates its quality to clients, investors and supervisory authorities.
The service is being launched through the company CertX, in which the reinsurer holds a majority stake.
Fabian Winter, chief data officer at Munich Re, said: “CertAI is a service that will give Munich Re yet another opportunity to contribute its qualified risk perspective to a forward-looking technology at the earliest possible stage and on the broadest scale. CertAI will enable us to address the significantly expanding coverage needs of AI solution providers in the market. At the same time, it will help us to hone our analytical skills in assessing the use of artificial intelligence for the insurance risks of tomorrow.”
The CertAI risk service is the second AI product Munich Re has cooperated in, after aiSure. Both initiatives are the result of Munich Re’s global innovation activities and were developed in the company’s own venture building and incubation programme. Whilst CertAI can help to define whether AI systems meet the requirements regarding trustworthy AI, Munich Re said aiSure focuses on the performance of these systems and covers the financial costs resulting from the reduced performance of an AI application.
Jens Henkner, director of CertX, said: “CertAI supplements our offering, and we, in turn, are contributing our own know-how to it. We look forward to engaging in a long-term partnership which Munich Re has underpinned with its strategic shareholding. Going forward, I am certain that we will be able to explore more global business ideas together that combine certification and insurance.”
