AI to enable more proactive and predictive risk management

Advanced analytics, powered by artificial intelligence (AI), is set to transform risk management, helping companies predict future risk and automate aspects of risk identification and mitigation, experts have told Commercial Risk Europe.

Demand for risk data is growing but so too is the availability of data and the tools to collate and analyse information, according to Jim Wetekamp, chief executive officer of Riskonnect.

The need for more and broader risk information is increasing as companies seek to understand risk beyond insurance, looking at a wider range of risks, a longer horizon and across more scenarios throughout the value chain, he explained. However, the volume of data and complexity of the task increasingly requires more powerful tools, he added.

AI, in particular, will enable risk managers and insurers to collect and analyse data quicker and more effectively, according to Wetekamp. “You are seeing AI change work and communication today. Where you have an opportunity for AI in risk management, and where it is potentially disruptive, is helping companies understand the risks to strategy over a longer time horizon. AI has huge potential to make companies more agile and faster to adapt,” he said.

“What is new, and changing, is where risk has to be in the conversation around the need for companies to be more adaptive and faster to change to market opportunities,” he added.

AI can help automate and enhance tasks involving large volumes of data, such as sorting and summarising, freeing up resources. It also has the potential to forecast outcomes, provide guidance and help make more informed decisions, explained Jörg Bertogg, chief operating officer for commercial insurance at Zurich Insurance.

“In the context of risk management, AI may be able to support with identifying emerging trends that might be missed by humans or more traditional methods of analysis, as well as suggesting preventive measures, effectively transforming risk management processes into being more proactive,” he said.

Marsh has seen increasing interest and investment in data analytics from risk managers, according to Brad Saunders, analytics development leader at the broker. It is working with more and more companies to help them use data analytics to inform their risk and insurance decision making, he said.

“In the challenging market there has been an increased focus on the value and efficiency of risk transfer and financing, and many are taking a more scientific view on how they buy insurance as they look to manage their insurance costs and rethink their self-insured retentions,” Saunders said.

“And with Covid and geopolitical risk, supply chain disruption etc, many companies are also looking to make more predictive and strategic decisions, and evidence and quantify value, supported by data analytics. These are certainly interesting times,” he said.

According to Saunders, risk managers are currently using data analytics to drive efficiency and gain insights. Companies are using machine learning and analytics to collect data, identify trends in claims, and inform risk transfer and financing. Clients with more sophisticated risk management are increasingly using advanced predictive analytics to understand and quantify risk and inform strategic decision making, he said.

Risk and insurance is, however, only now scratching the surface of what can be done with advanced analytics and AI, according to Wetekamp. “It is currently limited by the access of data needed to feed it, and the accuracy of that data. But this is also where AI can help, getting the right data to drive the analytics,” he said.

McKinsey estimates the total potential value of analytics to the insurance sector at €1.2trn but that only a fraction of that potential value has been unlocked.

“We are indeed still in early days of uncovering the full potential of AI and data analytics for risk management and insurance. Their potential to change both risk assessment and transfer is vast, with implications for how companies approach these processes,” said Bertogg.

“These technologies have a potential to lead to more proactive risk management, more accurate pricing, and more informed insurance buying, marking a transformation in how businesses perceive and navigate risk,” he said.

AI will help take data analytics to the next level, according to Wetekamp. “AI will really change scenario analysis and planning. The ability to analyse low probability high impact scenarios is currently limited by people. But AI can cover more scenarios and think of ones you have not thought of. Right now, it is very human dependent,” he said.

AI and analytics can continually monitor and assess compliance, including for risk management, explained Wetekamp. “AI can help capture information on the controls, policies, behaviours etc that are put in place in the business for risk mitigation. It can tell you whether the protection put in place to mitigate risk are working, or when you need to take decisions before the risk has manifested,” he said.

The technology can also help risk managers direct risk management resources and investment more effectively and efficiently, from enterprise risk through to insurance, said Wetekamp. “AI can help understand how much to spend at each layer and whether it is working,” he said.

The combination of more internal and external AI sophistication and improved computing power is enabling deeper insights into risk, more accurate policy pricing, proactive risk mitigation, streamlined claims processing and enhanced customer experiences, all driven by data, explained Bertogg.

“This is a further evolution towards a more data-centric model in the commercial insurance industry. At Zurich, AI-powered automation already helps us create a more simplified and insightful process and experience for our customers and brokers, which enables us to proactively provide information for them to validate and generate actionable insights rather than sending countless requests for information,” he said.

The advance of data analytics in insurance and risk management is fuelled by growing data availability through IoT and Industry 4.0, as well as the ongoing evolution of AI and machine learning technologies, said Bertogg.

“These elements provide a wealth of actionable information for risk assessment and management. Simultaneously, the accessibility and capability of data analytics tools are improving, enabled by user-friendly, cloud-based platforms, allowing even small to medium-sized businesses to harness these sophisticated analytics resources,” he said.

Tech trends such as wider IoT usage, sophisticated cloud services and AI evolution are accelerating the transformation in insurance, explained Bertogg. Specifically, AI’s advanced pattern detection and predictive capabilities may enable greater depth in risk assessment.

“This may allow insurers, for the benefit of the customers, to customise coverages even more, foresee potential risks, streamline risk mitigation and quicken claims handling, as some of the examples of a shift towards a more pre-emptive and tailored approach in insurance,” he said.

However, with all these advancements, there will also be new challenges, added Bertogg. “Issues related to data privacy, ethical use of AI and the need for new skill sets will have to be given greater consideration. Moreover, it will be important for insurers to maintain the human touch in their interactions with customers, ensuring that technology enhances rather than replaces personal service,” he said.