By Tendai Makaripe
The rise in prominence of Artificial Intelligence (AI) tools has revolutionised the socio-economic and political lives of people across the globe.
Artificial intelligence is the simulation of human intelligence processes by machines, especially computer systems.
Specific applications of AI include expert systems, natural language processing, speech recognition, and machine vision.
In general, AI systems work by analysing large amounts of labelled training data for correlations and patterns. These patterns are then used to make predictions about future states.
For example, a chatbot that is fed examples of text can learn to generate lifelike exchanges with people. Similarly, an image recognition tool can learn to identify and describe objects in images by reviewing millions of examples.
New generative AI techniques can create realistic text, images, music, and other media.
The insurance sector has been resistant to change for centuries but the AI revolution has caught up with it too.
Insurance is quickly shifting from its current state of “detect and repair” to “predict and prevent,” transforming every aspect of the industry in the process. The pace of change is also likely to accelerate as brokers, consumers, financial intermediaries, insurers, and suppliers become more adept at using advanced technologies to enhance decision-making and productivity, lower costs, and optimize the customer experience.
The use of AI in Zimbabwe can have numerous benefits for the industry and its clients. Fraud is a major concern for insurance companies, and AI is a key watchdog in the fight against fraudulent claims.
Insurance fraud refers to intentional dishonesty directed at or by an insurance firm or agent, aimed at obtaining financial advantage.
This fraud can occur at various stages involving applicants, policyholders, individuals filing third-party claims, or professionals rendering services to those claimants.
Likewise, insurance representatives and company employees may also engage in fraudulent practices. Widespread types of fraud include inflating claims, also known as “padding”, distorting facts in insurance applications, making false claims for non-existent injuries or damages, and contriving accidents.
The insurance industry in Zimbabwe is particularly plagued by fraudulent claims, over-insurance, and the proliferation of sham insurance companies.
Addressing journalists at a recently held insurance and pensions mentorship program organised by the Insurance and Pensions Commission (IPEC) and National social security authority (NSSA), Director Insurance and microinsurance at IPEC, Sibongile Siwela said: “The insurer and the insured must act in good faith towards each other. The duty of utmost good faith requires an insured person to disclose all material information with regards to the subject matter of insurance relevant to the insurer’s decision to accept or reject the risk…”
However, some have been acting fraudulently.
Research conducted by Fainos Chinjova and Madeline Mujakati in their study “An Assessment of the Impact of Innovation on Insurance Fraud Management in Zimbabwe” reveals that insurance fraud has led to a surge in costs by 158 percent, escalating from US$310 million to US$776 million between 2018 and 2019.
Academic Michael Porter notes that costs incurred by companies due to insurance fraud affect the competitiveness of the organization. AI can help fight insurance fraud. Samsung notes in a blog post about insurance fraud prevention that it is all about detecting patterns that might escape human cognition. Research has shown that through machine learning algorithms, AI can be trained to recognize patterns of fraudulent behaviour over thousands of claim records, identifying indicators that would be impossible for a human to spot.
AI can analyze huge datasets from different sources to identify anomalies, suspicious patterns, and correlations that might indicate fraud.
“AI can process claims in real-time, flagging potentially fraudulent claims as they come in. This can prevent fraudulent claims from being paid out, reducing the financial impact on the insurance company. With AI, fraudulent images submitted as proof during the claims process can be detected,” said insurance analyst Innocent Tinarwo.
“For example, AI can identify if an image has been tampered with or if it has been used in a previous claim.”
French AI start-up firm Shift Technology incorporates this technology in their fraud prevention services, which have already processed over 77 million claims.
Samsung writes in a blog post that the cognitive machine learning algorithms have reached a 75% accuracy rate for detecting fraudulent insurance claims.
The ML algorithms provide details on suspicious claims with potential liability and repair cost assessments and suggest procedures that can resolve and enhance fraud protection.
Automation driven by AI not only makes the claims process more efficient but also reduces the potential for human error.
The distribution chain in the insurance industry is winding and complex. A series of middlemen examine information between the insured and the carrier, leading to a lot of human error and manual work that slows the process. However, AI is starting to fix that problem.
According to analyst Fortune Nyamusa: “Algorithms can reduce the time and number of errors as information is passed from one source to the next. By logging in to a portal and uploading a PDF, the insurer reduces the amount of data entry and re-entry and increases accuracy.”
AI can process vast amounts of data much faster and more accurately than humans. It reduces the chances of manual errors in data entry, interpretation, and processing, leading to more precise risk assessment, pricing, and claim processing.
Nigerian-based insurance expert Ekerete Ola Gam-Ikon noted in an interview with this publication that AI can assist in efficient claims processing.
“AI can help in automating damage assessment using image recognition technologies. It can evaluate photos of a damaged asset, estimate the cost of repair, and reduce errors in claims estimation. Underwriting and pricing can also be improved,” said Gam Ikom.
“AI can help in evaluating and pricing risk more accurately by considering a wider range of variables and their complex interactions that a human might overlook.”
Not only can AI technologies enhance operational efficiency and insights, but they can also facilitate the creation of novel solutions and insurance coverage for risks that were previously uninsurable.
Swiss Re has developed a parametric Flight Delay Compensation system, powered by an AI model capable of predicting flight delays. In the case of a delay, insured customers receive an immediate payout, eliminating the need for filing a claim.
This system utilizes over 200 million historical data elements, and the machine-learning capability of its pricing engine allows for rate adjustments based on data from more than 90,000 daily flights.
Moreover, the advent of chatbots serves as a significant AI contribution to improving customer service. These automated responders can guide customers through various inquiries without the need for human assistance and are available around the clock, which is not possible with human teams.
For instance, if a customer requires assistance to access their account, they can get immediate help from the chatbot directly on the insurer’s website. This feature could potentially resolve customer issues swiftly.
While AI can have positives in the local insurance sector, Zimbabwe is still lagging behind in embracing its use due to a myriad of factors.
Zimbabwe, like many developing countries, may not have the necessary technological infrastructure, such as stable internet connectivity, to fully support AI systems. Not all insurance customers or employees may have access to necessary digital devices, which would limit the effectiveness of AI-driven services. “Implementing AI systems requires highly skilled personnel, which might be in short supply in Zimbabwe. The lack of local AI expertise could hinder the development and maintenance of these systems and might necessitate dependence on foreign expertise, which could be expensive and potentially unsustainable,” said Tinarwo.
Algorithmic bias is a problem that is worth noting. If AI models are trained on biased data or not properly localized for Zimbabwe, they can result in unfair outcomes or decisions.
For instance, an AI model might inaccurately assess risk for certain demographics if it doesn’t have enough data on them. Zimbabwe also lacks a specific regulatory framework for AI.
The lack of regulations could lead to misuse of AI technologies or problematic implementations, and it may create uncertainty for companies wishing to invest in AI.
It is in this light that the Insurance and Pensions Council ought to take this idea on board and craft a framework that regulates the use of AI in the sector.
