As we know, Generative AI has evolved into a genuine disruptor, outpacing earlier technologies like the internet in adoption speed and investment levels. Enterprises today are at a pivotal moment: those that embrace powerful new forms of AI with urgency will gain a significant edge, while those that hesitate will struggle.
Despite its transformative potential, many companies in the insurance sector are still experimenting, doing very little, or stuck in “pilot purgatory”.
It’s time to move beyond dabbling and embrace what’s possible.
A small European insurance brokerage offers a compelling example of AI’s new powers. It leveraged ‘Agentic AI’ to replace its entire operations department with non-human workers. The results were dramatic: substantial cost reductions, enhanced productivity, and industry-beating profitability.
The case study we feature below is quite mind-blowing and demonstrates the ‘art of the possible’…today.
Agentic AI: The Next Frontier
Traditional AI solutions assist human employees to be more productive, by creating content, interacting with customers, providing data insights or automating routine processes.
In contrast, ‘Agentic AI’ makes autonomous decisions, executes complex tasks, adapts to changing conditions, and collaborates with humans and other AI systems—effectively acting as a non-human workforce that can supplement traditional teams and augment the whole organisation.
Imagine entire business functions run by intelligent agents—not just chatbots that follow scripts, but autonomous entities capable of nuanced problem-solving and coordination. This new frontier is here today.
Case Study: When Necessity Drives Innovation
A small European insurance brokerage found itself with a very depleted operations team. Faced with this crisis, the CEO sought advice from an entrepreneurial friend: could AI replenish or even replace the whole organisational unit?
The answer was yes.
The entrepreneur—let’s call him Jim—designed an operations department composed entirely of AI agents. This new ‘team’ was made up of entirely non-human digital workers covering functions such as underwriting, finance, customer service, policy management and IT. These agents worked autonomously, problem-solving and innovating to achieve shared objectives, communicating over Slack. A ‘manager bot’ was developed to coordinate the team’s activities. Jim – the only human in the loop – oversaw the new team.
Within a few months, this AI team had surpassed the previous human team’s efficiency. The claims ratio, a key profitability metric, dropped dramatically from the previous 60%, transforming underwriting profit. The cost of running this agentic team was a fraction of a human team, creating an even bigger impact on net profitability. And, the bots never took time off for lunch, travel to work or to sleep – they worked continually constantly to develop, test and iterate new approaches to achieve the objectives that were set for them.
However, the AI agents were so effective that it quickly raised ethical issues: they were too ‘innovative’ at reducing claims ratios.
The CEO realised it was crucial to balance profitability with social responsibility. Jim had to train the bots to align with company values and regulatory expectations, ensuring that AI outcomes served both profitability and societal good.
(You can read the full case study here )
Hierarchy of GenAI-Enabled Applications
What I’ve just described is probably the most advanced deployment of ‘Agentic AI’ anywhere in the insurance industry today.
When I first shared it, many people thought is was made up. But it’s not – it’s a true story, using low cost tools that were readily available last year.
Its implications are very profound. Organisations can improve customer experience while reducing operational costs. They can scale operations without proportional increases in staffing. Traditional insurance sector margins can be significantly improved through more agile targeting, pricing, underwriting, claims handling and overall process efficiency.
To help leaders appreciate the full range of AI opportunities, and where ‘Agentic AI’ fits in with other forms of GenAI-enabled systems, we’ve created a strategic framework called ‘The 9 types of Organisational Augmentation’, per the diagram below.
Here’s a definition of each type of GenAI-enabled augmentation system, plus an equivalent roles that human employees have typically been responsible for.
| Type | Definition | Human Equivalents |
| Content Generators | AI systems that generate, edit, and optimize various forms of content | Policy Documentation Writer |
| Dynamic Responders | AI systems that enable dynamic, context-aware conversations and interactions. | Policy Inquiry Specialist |
| Digital Assistants | AI systems that support daily operational tasks and workflow management. | Claims Processing Coordinator |
| Decision Support Analysts | AI systems that enhance decision-making through analysis and recommendations. | Actuarial Analyst |
| Deep Domain Specialists | AI systems with deep specialization in specific professional domains. | Specialty Risk Assessor |
| Cross-Domain Connectors | AI systems that integrate expertise across multiple related domains. | Insurance Broker |
| Professional Collaborators | AI systems that actively collaborate with and enhance professional capabilities. | Risk Engineering Consultant |
| Autonomous Agents | AI systems capable of independent decision-making and action within defined parameters. | Chief Underwriter |
| Multi-Agent Teams | AI systems that function as coordinated teams handling complex organisational functions. | Complete Claims Unit |
Each of the nine distinct types of GenAI-enabled systems offer unique capabilities and strategic value. Rather than representing sequential maturity levels, these types represent different ways organisations can enhance their capabilities through AI. Organisations can—and often should—implement multiple types simultaneously, depending on their strategic needs, readiness, and objectives.
The potential benefits of Agentic AI, in green in the diagram, are significant. Organisations can improve customer experience while reducing operational costs. They can scale operations without proportional increases in staffing.
While employees can be more productive when augmented by AI capabilities across the whole spectrum of our framework, the challenges are equally substantial. Data quality and integration issues can undermine effectiveness. Security and compliance concerns must be carefully managed. Employee resistance may emerge if the transformation isn’t handled sensitively. There’s also a risk of over-dependence on specific vendors or technical approaches.
Issues for Insurance Leaders to consider
- Advanced AI and New Forms of Automation Are Here: The tools Jim used—existing language models combined with other technologies—are available today. Agentic AI can replace traditional human roles, handling routine tasks, making decisions, and solving problems. In doing so it augments the capabilities of the whole organisation.
- Productivity Gains: AI agents can surpass human teams, optimising processes and delivering outcomes beyond traditional capabilities.
- Cost Reduction and Scalability: Eliminating additional human-related costs enables businesses to reinvest in growth or simply enjoy enhanced profitability. AI agents can scale without proportional increases in cost.
- Ethical Challenges: AI’s ability to optimise narrow goals can lead to ethical issues if unchecked. Leaders must ensure that AI aligns with company values and broader societal responsibilities.
Act Now
The rapid adoption of AI means those in the insurance sector must act now to maintain their competitive position. Waiting, or dabbling, is a high-risk strategy. Companies using AI effectively will take market share from those who don’t. Enterprises need to decide which functions to augment and which to automate—and do so quickly.
From Experimentation to Execution
The primary reason many enterprises are stuck in pilot projects is the lack of a coherent AI strategy. Moving to real-world implementation requires a framework that aligns AI initiatives with business goals.
Consider the ‘AI Value Creation Pyramid’ to systematically approach AI deployment. By identifying impactful opportunities and establishing governance for ethical use, companies can unlock AI’s full value. It is not enough to “do AI”; enterprises must “do AI strategically” to maximize impact.
The Call to Leadership
Leadership is critical for AI adoption. The insurance brokerage succeeded because the CEO took a calculated risk and trusted Jim to design a new solution, navigating both technical and regulatory challenges.
Leaders must also ensure they attract the right talent—people like Jim, who have both technical expertise and industry knowledge.
The insurance brokerage case study underscores a truth about AI today: the future is already here, it’s just not evenly distributed. Enterprises have an opportunity to harness AI to transform operations, reduce costs, enhance productivity, and create new value propositions.However, this opportunity requires boldness, vision, and an immediate shift away from incrementalism. The question for enterprise leaders is not “if” AI should be adopted, but “how fast” and “where” to deploy it for the most significant impact. The time to act is now, and those who seize the moment will lead the insurance industry into the future.
About the author: Simon Torrance, Founder & CEO, AI Risk, is an entrepreneur and senior independent advisor to Boards and Leadership Teams on new growth strategy, technology innovation, systemic risk and venture building. In 2023 he ran the first in-depth think tank on AI Risk, supported by leading global organisations. Simon is also a member of the World Economic Forum‘s ‘Accelerating Digital Transformation’ executive working group, a guest lecturer at Singularity University, and co-author of ‘Fightback – how to win in the digital economy’.
