David Lynch, bolttech

Insurtech: 6 AI trends reshaping the industry

David Lynch, bolttech

February 25 2026

AI has become the defining technology of our time, entering the hands of everyday consumers and employees faster than any innovation in recent history.

The speed of adoption has been remarkable, mirrored by an equally intense wave of hype, but real business impacts are beginning to surface.

What was once the domain of specialised data teams is now integrated into day-to-day operations within insurance businesses.

As the hype continues, so does the imperative for insurers, insurtechs and distributors to look beyond the noise to the underlying forces at play.

Several powerful trends are already shaping the next chapter of AI – and with them, the future of how companies will build products, serve customers, and compete in the insurance and protection marketplace.

1. Conversational commerce becomes a new default

One of the most striking shifts is the rise of AI-led conversational commerce. The enormous investment poured into large language models (LLM) was always going to push developers to embed these capabilities into familiar applications and experiences.

That is exactly what has happened: everyday tools have quietly absorbed generative AI, and consumers now find themselves engaging in search, discovery and purchase journeys through conversational interfaces rather than clicks inside of traditional websites.

This shift is visible in insurance just as much as ecommerce. Product roadmaps that a year ago relied on fixed, linear customer journeys have been reimagined around real-time dialogue and dynamic customer engagement.

Our teams at bolttech in North America, for example, have spent months reshaping the way insurance is bought and sold through conversational pathways, embedding AI-driven dialogue and agent-led engagement into distribution journeys as customer expectations fundamentally shift.

Agentic commerce is here to take the time and pain out of product search, underpinned by Universal Commerce Protocol, the open framework allowing AI agents and platforms to discover, interact with, and complete purchases across merchants seamlessly, without custom integrations.

Global technology giants are responding to the trend. OpenAI has begun reintegrating its models into browsers and apps, while other providers are racing to launch AI-centric experiences that place conversation at the centre. What we are witnessing is a redefinition of how consumers engage with businesses. Slick user interfaces are being bypassed with AI that understands intent, remembers, and intelligently adapts.

2. The shift from SEO to GEO – Generative Engine Optimisation

For decades, businesses competed to be visible on Google’s search results page. But a structural shift is underway. As consumers turn to LLMs for answers, advice and product recommendations, companies must now optimise not only for search engines but also for generative engines.

This requires a completely different playbook. LLMs retrieve information, restructure it and present it back to consumers in ways that often bypass the need for consumers to visit traditional websites themselves.

Brands therefore need to understand how they are being represented inside these engines – how their products are surfaced, what information is being prioritised, and where inaccuracies or outdated content may be influencing customer decisions.

At bolttech, we have been making this shift as well, deriving significant insights on how these models operate. What we are seeing is a new frontier of digital strategy: companies adjusting campaigns, content assets and data structures to ensure they show up correctly inside AI models rather than simply climbing search rankings.

The behavioural signals only reinforce the speed at which this shift is taking place. Data suggests that engagement times within conversational AI are dramatically higher than on traditional search. Younger “AI natives” are accelerating the shift and defaulting to AI first for information or product recommendations.

3. Small language models making an impact

While LLMs dominate headlines, the rise of small language models (SLMs) may prove even more consequential for enterprise adoption. Many prompts, tasks and workflows do not require the breadth of a massive, general-purpose model. They require speed, specificity and cost efficiency – and that is where SLMs excel.

For insurance players in particular, these models can be trained with deep industry context, enabling them to understand the nuances of insurance language and regulatory terminology far more accurately than general models. They can also run on the edge, leveraging the AI chipsets in modern smartphones. That means near instant responses, minimal latency and almost zero marginal cost.

This matters because organisations everywhere are experimenting widely with AI, but very few are deploying it at scale. As usage grows, the economics of inference – the ongoing cost of running models to generate outputs – become critical. The difference in cost between model types is significant, and energy efficiency is becoming a strategic concern in its own right. With sustainability targets and the rising carbon footprint of AI compute, SLMs offer a path toward scalable, responsible deployment.

4. The robotic workforce expands as autonomy accelerates

AI-driven automation is not new, at least in earlier incarnations of far less intelligent solutions than we have today. We’ve been through decades of robotic automation dating back to the early 2000s, but the arrival of generative AI has dramatically expanded what automation can do for business. The length of tasks AI systems can perform autonomously is now doubling every month.

A year ago, autonomy plateaued at around four hours; today, Anthropic claims that models like Anthropic’s Claude Sonnet 4.5 can run independently for over 30 hours.

This shift is transforming internal operations. New AI-enhanced tools are giving teams capabilities that barely existed a year ago. The impact is not simply about productivity. It changes research is conducted, how problems are solved, how workflows are designed, how products are engineered and built and how teams think about where human judgment is most valuable. Whilst consumer-oriented humanoid robots are also on the horizon, let’s also consider they will have insurance and protection needs.

The single greatest unlock remains in the power of choosing the right use cases to invest in. The technology is ready. The tooling exists. What matters now is identifying the specific problems across the business that AI can create a meaningful impact. For example, at bolttech, we have deployed AI across the value chain, from accelerating claims intake and triage, to extracting insights from complex documents and enabling faster, more intuitive customer support through conversational AI. The opportunity is enormous, but it depends on the creativity and insight of teams within each organisation.

5. The arrival of frontier agents

Perhaps the most transformative trend is the emergence of agentic AI – systems that can plan, reason and execute tasks end-to-end with a level of autonomy previously unimaginable. Unlike classic agents, AWS announced in late 2025 that these more powerful frontier agents can run for days, maintain memory, operate in parallel and integrate with control frameworks that allow them to act within highly complex environments.

The major cloud providers – hyperscalers – such as AWS, Google Cloud and Microsoft, are making it possible for companies to easily build their own agents, tailored to specific functions.

Given the immense R&D needed to build more intelligent capable frontier agents, examples like AWS’s DevOps and security agents demonstrate how quickly this space is evolving. We have hundreds of agents already live within bolttech and see incredible potential to supercharge with these frontier agents. We envision that every single role will have some level of agentic support.

6. The silent backbone: Connected ecosystems and data

Amid the excitement, it is easy to overlook the quieter trend powering all of this: the rise of vast, connected ecosystems that provide the data that serve as fuel for AI. Homes today routinely contain dozens of connected devices. Modern EVs usually have over 100 different sensors. The number of connected devices is expected to climb towards 50 billion globally by 2030 and this is providing the fuel (data) for AI to do it magic. 1

These connected devices – from household IoT sensors and smart appliances to wearables, EV telematics and the growing range of in‑home robotics – generate the data streams that enable insurers to anticipate equipment failure, detect risk conditions earlier and intervene before loss occurs. These interconnected devices allow prediction and prevention at scale and with positive impact far beyond anything the insurance industry has been able to achieve before the advent of AI.

Entirely new insurable categories are emerging, including EV batteries, and household humanoid robots – a segment projected to reach 10% adoption within a few years and potentially become a US$5 trillion market by 20502. Protection services are evolving far beyond the financial risk transfer that insurance offers. Consumers are beginning to experience a world where many categories of claim and loss may become a thing of the past.

Together, these trends signal a future where AI becomes the operating fabric of insurance – powering smarter products, more proactive protection and entirely new ways to serve customers at scale.

This article is written by David Lynch, Group Chief Technology & Operating Officer, bolttech.  Lynch is based in Melbourne. 

MORE FROM: Comment
Partner Content
Join the mailing list Designed to help you stay ahead, our daily newsletter provides a round-up of the latest news and insights, delivered straight to your inbox.
Print Edition
Subscribe to receive our quarterly print magazine.