How to plan a clearly articulated generative AI strategy
Carriers still using legacy technology will struggle to leverage generative AI and other advanced technologies for their multiple use cases, including improving pricing, portfolio management, claims handling and customer service
The extended hard market has obscured the problem of both ‘classic’ and ‘modern’ legacy systems
Having shaken off inertia, overcome fear, secured the budget, held the tender, conducted the pilot and pushed through IT modernisation, the average insurer may want to rest on its laurels.
Unfortunately, that tech overhaul may well have created “modern legacy” systems that are ill equipped to respond to changing market dynamics or to exploit the opportunities presented by generative artificial intelligence (AI) and other advanced technologies.
“Legacy” as an adjective, whether uttered defensively by a chief executive on investors’ day or by an incoming chancellor of the Exchequer, is rarely a good thing. In the context of technology, you might assume it applies only to on-site mainframe systems hulking in the darkest recesses of dingy office blocks, costly to maintain and sometimes impossible to repair. But a technology stack does not have to originate in the era of the Dukes of Hazard to be outdated.
Legacy solutions may now include cloud-based systems that are as little as five years old. They may also include managed service platforms that require separate cloud support. These can create integration challenges, a particular problem for London market groups that need to connect to systems such as ECF and Xchanging.
The tide goes out
The extended hard market has obscured the problem of both “classic” legacy and modern legacy systems. Organisations had ample opportunity to make money, provided they exercised basic underwriting discipline. Some still operate with years-old legacy systems, with solutions siloed by function.
Complacency is partly to blame; however, the variation and complexity of the underlying systems (policy, underwriting, pricing, claims and so on) makes modernisation daunting. So too do the myriad forms information may take: Word documents, spreadsheets, PDFs, emails, phone calls, call logs, paper notifications and (incredibly) even faxes.
As the cycle continues to turn, the need to write three times the number of policies and assess three times the volume of submissions to stay in the same place means companies that have failed to update their technology will struggle compared with those that have stayed up to date
However, as the cycle continues to turn, the need to write three times the number of policies and assess three times the volume of submissions to stay in the same place means these companies will struggle compared with those that have stayed up to date. Leading insurers will harness generative AI to analyse submissions data at pace and assess harder-to-underwrite risk.
Carriers that have modernised since the turn of last decade have not invested in vain, provided they have good APIs and well-structured databases. However, they probably have not considered how to optimise this investment and leverage generative AI because the need has not arisen in the hard market. Many have also over-indexed on tech investment being primarily a driver of operational efficiencies rather than revenue growth. Generative AI can facilitate both and now is the time to sweat that earlier technology investment.
Clear roadmap
Modern carriers, brokers and managing general agents should take the next step by devising a clear and thoughtful roadmap to generate efficiencies and support growth through the application of generative AI. Key internal stakeholders need to be fully on board – and that means more than the chief technology officer and their team.
The age profile and professional background of C-suite insurance executives means they are probably not habitual early adopters but when they understand the possibilities of generative AI to spur growth and improve relevance, the conversation becomes easier.
Incorporating generative AI into technological systems also requires careful governance and astute change management. Some users like underwriters might be reluctant to deviate from traditional ways of working. However, the growing volume of data at our disposal has already long surpassed what humans can handle and a clearly articulated generative AI strategy should convince sceptics the technology will support rather than supplant them.
The consequence of inaction is severe: carriers still using legacy technology will struggle to leverage generative AI and other advanced technologies for their multiple use cases, including improving pricing, portfolio management, claims handling and customer service. They will find it harder to cut costs and create operational efficiencies and, to underwrite the business, they need to hold their ground as the market softens.
Warren Buffett famously said you only discover who has been swimming naked when the tide goes out. As rate growth continues to slow – or go into reverse – expect to see some red faces. Those who have relied on market dynamics to defer technological transformation will look completely exposed, while the coverage provided by the metaphorical sagging Speedos of their “modern legacy” peers will be limited. Neither is a good look.
Brandon Nuttall is chief digital and AI officer at Xceedance