top of page
  • Mayan Kansal

GenAI and how to think about its impact on Insurance

Why are we always late to the party?

Moving beyond the repetitive explanation of Generative AI (or its breakdown into Silicon, Foundational, Infrastructure, and Application Layers), let’s try to dive deep into what really matters to us its impact on the Insurance Industry. In recent years, the insurance industry has rather witnessed slower adoption of newer technologies. The insurance industry's cautious approach to the adoption of certain technologies may stem from these technologies’ tendency to overpromise and underdeliver. As a result, we have been hesitant to be early adopters of such technologies. GenAI certainly has the potential to impact the whole value chain of Insurance, from Customer Support to Claims, to Underwriting. However, by the end of this article, let’s try to build actionable frameworks on how to think & decide about GenAI’s adoption into our Insurance Businesses.  

Understanding the Wizardry of GenAI, in the Context of Insurance

Let’s keep it simple. What unique powers does GenAI have in its armour, that this time we’re more certain of being early adopters? Well, we’re in luck this time: (Large Language Models) LLMs are good at summarising vast amounts of complex information → exactly what Insurance Professionals do manually, all day long. In the insurance industry, GenAI can aid and work alongside human professionals across the value chain, to enhance efficiency, accuracy, and decision-making.

Impact in Insurance: and how to think about it

So basically, LLMs can Summarise, Create, enable Semantic Search, or act as Virtual Assistants. Let’s try to Map insurance use cases (not an exhaustive list) with these core capabilities of GenAI:

Use cases in Insurance, mapped with the fundamental GenAI capabilities

  • Summarisation: Summarise risks/claims, query data, run sentiment analysis- Creation: Customer Support/Sales Chatbots, Writing Assistance, Documenting & Reporting

  • AI Co-pilots: Co-pilots for Workflow Automation, and Decision Making (e.g. Claims, Support, Underwriting)

  • Semantic Search: Internal search engines (over semantic layers of pre-processed information)

The capabilities of GenAI in insurance can be enhanced by combining it with other technologies, like IDP. To design and construct cohesive systems of complex workflows, innovators need to think in terms of combining multiple technologies. Such solutions can then assist insurance professionals from internal use cases like risk assessment, or policy drafting, to customer-facing use cases like claim support. Let’s now look at some of these potential, comprehensive use cases, by the functions they could affect. 

Use cases in Insurance, and functions they affect

Assessing risks is a crucial part of underwriting in the insurance industry, especially in LoBs that require extensive data analysis before providing a quote for an RFQ. Underwriters can automate risk summarization with LLMs. By analyzing inbound data using data-backed logic, underwriters can identify relevant risk factors, and analyze historical data, market trends, and customer information to address RFQs.

Impact: GenAI can help insurance businesses improve their operational efficiency by (semi) automating underwriting processes and triage. By addressing more RFQs more efficiently, it can enable faster quote-to-bind, while also reducing the time required to evaluate risks. This can lead to overall faster TATs, and save Underwriters' hours of manual tasks, unlocking faster and more accurate decision-making. Additionally, GenAI can minimize (human) errors, which can further improve the efficiency of the insurance business.

GenAI can impact the whole claims cycle. For instance, it can automate claims processing, detecting fraudulent claims and expediting legitimate ones. In health, GenAI can analyze vast amounts of data, including medical records, accident reports, and policy details, to assess the validity of claims and ensure efficient and accurate processing.

Impact: Improvement in Net Promoter Score (NPS), quicker response time from First Notice of Loss (FNOL) to resolution process with increased accuracy, and granting of authority to claims managers for better decision-making.

GenAI can enhance customer experience (CSAT) by providing human-like responses to vast customer queries. GenAI can power internal Co-pilots to empower FLS executives, agents, support executives, or POSPs, to respond to customer queries faster, with high accuracy. Or it can be integrated as customer-facing chatbots, for pre-sales and post-sales use cases, adding customer delight.

Impact: Improved query resolution leads to higher customer satisfaction (CSAT) and agent empowerment through increased internal efficiencies.

Potential Challenges, and Decision-making Framework

The potential of GenAI is immense, and it is poised to transform the insurance industry. As a key decision-maker, it is important to consider both long and short-term factors when deciding which LLMs to implement. Additionally, we must decide whether to build in-house or leverage solutions from InsurTechs and how to align GenAI solutions with our organization's AI policy. Insurance is an industry based on the principle of trust. To maintain customer confidence, organizations must prioritize data privacy and comply with standards.

Factors that should govern the internal decision-making process to implement GenAI solutions within an org

  • Cost of going Wrong (use-case wise)

  • Regulatory Data & Encryption needs 

  • Generic Model vs. Domain-specific Models

  • Build (proprietary or fine-tune) vs. Buy (License) LLMs

  • Core vs. Non-core Operations use cases

  • Time-to-market

  • Flexibility and Customization

  • Ongoing maintenance and updates

  • Cost and resource availability

  • Scalability

  • Integration with existing systems

There are various scenarios where insurance businesses can benefit from implementing GenAI-based solutions. However, a "one solution fits all" approach wouldn't work here. For instance, if the goal is to differentiate quickly, then faster time-to-market should be prioritized in customer-facing use cases. In such cases, licensing GenAI tools could be a better economic choice to validate use cases without wasting internal resources. Alternatively, collaborating with InsurTechs could be considered to lead internal GenAI use cases, which can free organizations from the complications of dealing with multiple LLMs. If an organization has in-house resources, it could speed up its time-to-market by collaborating with InsurTechs. On the other hand, organizations that are exploring straightforward GenAI capabilities might be better off leveraging off-the-shelf foundational model APIs. In cases where efficiency is a priority over differentiation, it might be better to offload it to GenAI-first InsurTechs. Relying entirely on in-house resources could be a costly mistake since outside domain experts can bring innovations.

Conclusion: don’t be late to the party

With various GenAI use cases for Insurance, one simple plan of action could be to be quick to experiment, validate, adapt, critique, rethink, or reject. Generative AI is here to stay and to only get better. Technology companies like are enabling faster adoption of GenAI solutions into Insurance businesses, by developing solutions aligned with unique domain-specific challenges & requirements like data security, and customisations. Insurance businesses need to move fast, to create their long-term differentiation strategy, and have solid data (from experiments) to drive decisions around the adoption of GenAI for various use cases, ultimately targeting scalable, compliant implementations across their organisations. (This article was originally published in India InsurTech Association (IIA) as a Thought Leadership article.)

14 views1 comment

1 Comment

Rated 0 out of 5 stars.
No ratings yet

Add a rating
2 days ago
Rated 5 out of 5 stars.


bottom of page