The AI Revolution in Insurance
Marc SEVESTRE
Senior Business Advisor X-PM | International Board Member | Former European CEO Global Insurance Company | Entrepreneur & Investor | Executive Coach
The discussions on Generative Artificial Intelligence (AI) have been dominating the media space since the arrival of ChatGPT (which dates back only to late November 2022). Not a sector exists that does not ponder, wish, or fear that its activities are on the verge of being completely impacted, revolutionized by the deployment of AI. We are only at the beginning, and the funds raised around the most promising projects and teams — at least on paper — are reaching dizzying heights. This enthusiasm is reminiscent of the dot-com bubble that eventually burst at the turn of the millennium (March 2000). With the launch of GPT-4o in May (o for omni, as in omni-channel), media buzz has further intensified. This was especially notable when actress Scarlett Johansson expressed anger upon discovering that one of the voices used by this "conversational" version of GPT was very similar to her own. Beyond the hype, let's project ourselves into the short to medium-term future (before the end of the decade) to consider the impact AI will have on our daily lives. Every sector will be transformed. To better understand the depth of the upcoming changes, let's focus on use cases related to the insurance sector.
?
AI and Insurance
Before asking ChatGPT for its own view on the subject, let's draw on our experience as the CEO of insurance groups at the European level to review the various activities that AI is expected to impact the most.
Generative AI allows to produce content (texts, images, conversations) similar to what a human can do, but automatically, immediately, and systematically. Schematically and intuitively, artificial intelligence can therefore be applied, albeit differently, to all insurance functions and therefore to the main ones. These include marketing, subscription including pricing, claims/compensation, and finally client management/relations. Without delving into the details of each activity, let's briefly analyze the impact that artificial intelligence should have — and the revolution is already underway — on these different functions.
AI and Marketing
Analyzing each client's specific situation, the evolution of their needs over time, optimizing their insurance coverage by best combining different products requires integrating a very large number of parameters. Each client is thus a unique case. Therefore, an undifferentiated offer, relying on a few typical profiles and pre-packaged solutions, does not meet the needs of each insured person. The analytical capacity of artificial intelligence will therefore identify the most appropriate response, integrating all the data collected prior to the recommendation.
This use of AI also helps to meet the regulator's demands for customer knowledge and the duty of advice that applies to insurance product distributors.
AI and Underwriting
Based on a detailed understanding of the client, their needs, and risk profile, AI highlights a differentiated offer that meets the client's expectations and is priced fairly. Certainly, the basic mechanism of insurance is the mutualization of ?risks. This means that "good" risks contribute to financing "bad" risks. Although insurers claim that there are no bad risks, only poor pricing, the right price is based on a form of pooling within the same category of insured. The probability of the event covered occurring varies among these different "clusters"; the riskier ones are assigned a higher tariff grid.
AI enables the precise determination of the risk profile of each insured person (life insurance) or each event (damage insurance) and thus proposes the most relevant pricing, the fair price.
AI and Claims
When a claim is filed with the insurer, they ensure that the request is legitimate. False statements on subscription and insurance fraud to claim undue sums are a constant concern and a sensitive issue that insurers must pay particular attention to. It is necessary to know where to place the cursor between a slow and meticulous examination that does not sit well with insured persons and a judicious payment. Overpaying affects accounts, which is reflected in the mass of insured persons by raising rates in the event of deteriorating technical results.
AI proves to be a valuable ally in tracking false statements, which has not escaped the tax authorities in their hunt for fraudsters. Crossing data from numerous sources, conducting cross-references, tracking multi-declarations, noting inconsistencies, AI allows controls to be pushed far beyond what the most astute detectives could identify. A well-known example, among others: An unscrupulous insured who declared sick leave with the complicity of a rogue doctor posted photos of vacations on a paradisiacal beach. AI can analyze the photos, link them to the numerous sick leaves issued by the healthcare professional, and unmask the imaginary sick man.
AI and Client Relationships
Management, back-office, the "bilge" for many insurers. A phenomenal number of thankless tasks, mountains of documents. But also numerous interactions with clients. Every insurer's dream is to make this process smooth and enjoyable for policyholders (the famous UX, user experience) with the use of natural language[1] chatbots, extremely automated to limit the financial burden of personnel. The Holy Grail is to transform this enormous cost center into a profit center and to take advantage of this interaction with the clients to review their situation, propose adjustments, introduce a more relevant offer given their evolving life situation over the years. Leave the tasks of administration to the machine. Retain humans for the most value-added part, providing policyholders with a quality customer experience and thereby keeping them loyal to their insurer and increasing the number of products and services held (cross-selling).
What better ally than AI, capable of organizing client data, while analyzing the entire situation of the policyholder and in view of existing and future needs to offer them a product portfolio optimization?
Through these examples that illustrate the current and future use of AI in insurance, we see how it revolutionizes this sector, as well as many others, by bringing major innovations and improvements.
领英推荐
Copilot's Perspective
So, why not ask Copilot how it sees AI transforming insurance? Here is its response:
? Risk assessment and fraud detection: AI's intelligent algorithms quickly analyze data to assess risks and detect frauds. This allows insurance companies to accurately identify client needs.
? Automated claims processing: AI facilitates automated processing of claims by analyzing information on incidents. This enables effective assessment of damages and precise estimation of compensation costs.
? Policy personalization: Thanks to AI, insurers can customize policies according to each client's specific needs.
? Improved customer service: AI-based chatbots and virtual assistants provide quick and accurate responses to client queries, thereby enhancing the customer experience.
In summary, AI is profoundly transforming the insurance industry by automating complex tasks, improving decision-making, and personalizing the customer experience.
Impressive! We had of course drafted our perception before asking Copilot for its input, and we did not alter it once we received its response. Our analyses converge on many points, even though Copilot, unlike us, does not have over thirty years of experience leading major insurance companies. This touches on the fears expressed by many professionals, senior executives, and artists ?of being replaced by machines. A massive destruction of qualified jobs, as we presented in an article on this topic: Chat-GPT, serious threats to white-collar workers[2].
"Generative AI is taking over areas that seemed uniquely human, such as reflection, creation, and intuition. This could mean the disappearance of three hundred million jobs, primarily cognitive tasks. The most affected will be executives."
This poses a serious threat to a wide range of functions with strong intellectual or creative content.
?
Threat or opportunity. Only time will tell. Depending on how we handle it, AI will be either an ally or an enemy. One thing is certain: for a senior executive, it is impossible not to integrate AI into their development plans, development strategy, and growth objectives. Massive investments are already dedicated to AI, and the biggest is yet to come. The way companies operate and their organization will be profoundly impacted. AI will reshape our world, perhaps even more so than the steam engine, electricity, or the internet did in their time.
Marc SEVESTRE (former CEO of AIG Life Companies for Western Europe and MetLife for Western Europe) and Copilot
#AI #ArtificialIntelligence #Chatbot #Chat-GPT #Copilot #Generative #Insurance #OpenAI
[1] ChatGPT relies on a Large Language Model (LLM) built on a deep learning algorithm, capable of performing various tasks related to natural language processing. The LLM is constructed on a deep neural network architecture based on "transformer" models. In essence, this allows for processing large volumes of data contextually. Different layers of neural networks combine to process the input text and generate the output content.
?
Director & General Secretary
5 个月How are you?