AI and the Future of Insurance: Transformative Potential or Incremental Change?
The influence of artificial intelligence (AI) on the insurance industry is undeniable, yet discussions surrounding true digital transformation have stalled. This situation raises a critical question: Are we genuinely on the brink of a revolution, or merely witnessing an enhancement of existing processes?
Efficiency or Innovation: What’s Driving Transformation?
When we talk about digital transformation in insurance, we often frame it in terms of efficiencies, automation, and the phasing out of outdated legacy systems. While AI contributes significantly to these areas, some argue it merely accelerates existing practices without ushering in groundbreaking changes.
Conversely, if transformation is about fundamentally reengineering how businesses function, engaging customers, and innovating products, then the focus on generative AI and large language models might be more suitable for keynote speeches than actionable strategies. It appears that many in the insurance sector have yet to rethink their collaborative strategies, particularly in how insurance firms partner with banks and agents to distribute products effectively.
Such sentiments were echoed at a recent industry event on digital insurance and captured in comments by DigFin.
The Growing Sense of Urgency
A dramatic shift seems to be confronting insurance executives. The inherently accessible nature of generative AI has raised alarms among C-suite leaders, signifying that this is no longer just a conversation about growth—it’s about survival.
As Orchis Li, General Manager at Gen Re Hong Kong, stated, “It’s no longer about digital transformation; it’s now about survival.” This urgency is pushing firms to adopt automation tools enthusiastically. Rather than being seen as top-down initiatives helmed by CTOs, genAI is now fostering projects driven at all levels, which many find invigorating.
Firms have increasingly digitized their operations, converting analog records into machine-readable formats. As a result, automation is weaving into the fabric of decision-making processes by creating seamless data flows.
A Surging Wave of Digitalization
Jim Qin, CEO of Zurich Hong Kong, highlighted his organization’s significant leap from just 25% to over 90% digitalization in a few short years. This transition establishes a robust foundation for data-driven decision-making: “Every piece of data is digital and structured, so we can use it,” Qin elaborated.
The next logical phase involves leveraging AI to unlock the full potential of this wealth of data, leading to improved automation in claims processing, underwriting, and customer profiling. The ultimate goal? An organization transformed into what can be described as “AI-native.”
Reality Check: The Rhetoric vs. the Ground Reality
However, not everyone is convinced that the current wave of AI advancements constitutes genuine transformation. The term “AI-native” can sound grandiose, and many are questioning its validity.
Indeed, while automation may be a perfect fit for many industry functions, the applications seen thus far—like predictive modeling during underwriting and claims—are merely refinements of existing processes. New efficiencies are welcomed, but merely improving established models does not constitute a transformative leap. This begs the question: Are insurers truly becoming “AI native?”
The Role of Causal Reasoning in Shaping the Future
According to David Piesse, a notable figure at the International Insurance Society, substantive advancements in the sector depend on the adoption of agentic AI—systems integrating large language models, causal reasoning, and human insight.
The crux lies in the term "causal." Today’s AI largely operates on massive datasets, identifying correlations rather than understanding causation. An AI can recognize that X is related to Y, leading to Z, but it often cannot ascertain the underlying reasons for these relationships. This lack of causal understanding can be detrimental, particularly in critical industries like insurance.
Despite the focus on achieving outcomes, the evident risks associated with AI-generated correlations shouldn’t be overlooked. These models, as illustrated by the rise of genAI, can sometimes yield inaccurate results, making dubious connections that could lead to errors.
Rethinking Business Models Amidst Transformation
Indeed, genuine digital transformation redefines not only the use of AI but the entire approach to business models. According to Piesse, the industry may currently be suffering from an “overdose” of genAI, suggesting that a deeper, more expansive reimagining is essential—particularly as we explore concepts like parametric insurance.
Navigating the Legacy Technology Trap
To achieve true transformation, firms must address the enduring presence of legacy technology. Overhauling outdated systems, some of which are decades old, can take years or even decades. The journey from traditional mainframes to contemporary, cloud-driven solutions is fraught with challenges.
Data integration also poses significant hurdles. While banks are making strides in this area, many insurance organizations face difficulties in consolidating their fragmented systems, hampering their ability to derive meaningful insights from data.
The Importance of Quality Data
In discussions about AI’s future in insurance, the quality of data cannot be overstated. As Gary Ho, CIO of AXA Hong Kong and Macau, pointed out, merely discussing AI is futile unless accompanied by high-quality data. The focus should lie not just in improving internal efficiencies but also in delivering value to customers and distribution partners through optimized processes.
Data integrity is a vital concern, particularly in AI-driven decision-making; ensuring that data is accurate, traceable, and secure from manipulation becomes increasingly important as reliance on AI grows.
Regulatory Hurdles: A Barrier to Innovation
Adding to the complexity, regulatory requirements often slow innovation. Selina Lau, CEO of the Hong Kong Federation of Insurers, explained that risk-based capital considerations and new accounting standards can consume resources and slow down progress. Legal and compliance teams wield considerable influence, and without their support, even well-intentioned initiatives can falter. “If the CEO says yes but legal/compliance says no, it’s a no-go,” Lau stated, emphasizing the regulatory nature of the industry.
Harnessing AI for Meaningful Impact
Despite these challenges, insurers are actively finding practical applications for AI. Many are experiencing significant efficiency gains through automation, allocating more resources to growth-oriented initiatives. For instance, Michael Shin, CEO of RGA Korea, described how AI-driven OCR has streamlined medical underwriting, significantly reducing errors and turnaround times.
More than just operational benefits, AI is paving the way for innovative products and market expansion. In healthcare, for example, advanced analytics can identify early indicators of conditions like Alzheimer’s, allowing insurers to create specialized offerings focused on early interventions.
A New Era for Insurance Offerings
Moreover, models such as embedded insurance and insurance-linked securities depend on real-time data accuracy and advanced risk assessment—particular strengths of AI. Such innovations enable insurers to provide tailored microinsurance and parametric products that can serve underserved markets, widening their reach and impact.
These emerging solutions represent nascent but promising developments, indicating a shift toward innovative business practices fueled by reliable data sources and computational power.
Leverage AI to Reshape Legacy Systems
Ultimately, the current phase of AI presents an invaluable opportunity for the insurance sector. Not only can it enable firms to shed their reliance on outdated mainframe systems, but it can also facilitate the recoding of legacy software to adapt to modern needs.
As the pace of change quickens, insights from leaders like Maxim Afanasyev, head of financial services at Google Cloud, reveal a growing sentiment within the sector. He notes that 37% of financial institutions have already implemented generative AI, with an additional 48% in the process of doing so. The insurance industry, in particular, is recognizing the critical need for AI-driven solutions, positioning itself at the forefront of this digital evolution.
In conclusion, while the insurance industry is at a pivotal moment, it is crucial to distinguish between superficial enhancements and transformative efforts. As firms grapple with the challenges and opportunities presented by AI, the journey towards genuine digital transformation is ongoing. The path ahead may be complex, but the potential rewards—improved efficiency, enhanced customer value, and innovative product offerings—are worth pursuing. This continued dialogue around AI and digital transformation is sure to persist in industry conferences and strategic discussions moving forward.
