The AI Book

CHAPTER 27Cutting to the Chase: Mapping AI to the Real-World Insurance Value Chain

By Nigel Walsh¹ and and Mike Taylor²

¹Partner, Deloitte

²Director, Deloitte

AI’s (re)arrival is hailed by some to be the fourth industrial revolution, and its impact on insurance is no exception. In this chapter we don’t attempt to reach consensus on the various definitions of AI. Instead, we seek to explain the applications of AI in insurance through the various layers of AI applications and logic, as illustrated in Figure 27.1. There are a number of functional capabilities that AI can deliver that mimics, or augments, human capabilities:

1. Computer Vision: analysis that enables the classification and tagging of objects within a video, image or any sensor reading. An example of this is the identification of an individual from a picture of their face or facial recognition.
2. Natural Language Processing (NLP): this encompasses both the understanding of written text and the generation of new text, for example, the understanding of questions and generation of text responses within a chatbot.
3. Voice: the transcription, enrichment and analysis of audio data, invariably resulting in NLP but with the additional information available during speech such as emotional tone.¹ An example of this is the augmentation of a chatbot to support spoken questions and responses.
4. Internet of Things (IoT): processing of any sensor information from a network of devices connected through the internet. This category might ...