Insurance has never been short on data. The challenge has always been turning that data into decisions that move the business forward. In a recent InsureTV discussion, Tony Shek, Global Insurance Lead at AutoRek, and Dorota Zimnoch, Global Industry Strategist at Microsoft, unpacked what separates insurers who experiment with AI from those who scale it into real, measurable outcomes.
Here’s what stood out.
Data is only an asset when it’s connected
Insurance is one of the most sophisticated operations in the economy. Claims, capital management, risk modelling, actuarial underwriting and pricing all run in parallel, and most insurers are still doing it across multiple legacy systems and fragmented data sets.
The insurers pulling ahead are the ones treating connectivity and automation as the foundation, not an afterthought. Connecting data holds real value, not just holding it.
From reactive reporting to real-time decisions
For years, insurance data was used to look backwards: collect the information, then underwrite, assess and price later before coming back with a decision. AI is flipping that model.
With more unstructured data now being brought into structured, usable form, insurers can embed data directly into live processes. That means decisions made on the spot with the customer, rather than days later. Data stops being a record of what happened and becomes a decision-making capability.
What connected data actually delivers
When insurers connect their data flows, the benefits show up in three clear areas:
- Efficiency: Many insurers hit scalability problems when the business grows faster than they can hire. Automation lets teams grow moderately while sitting in the driver’s seat, monitoring processes and focusing on higher-value work. The result is better data accuracy and faster, more informed decisions.
- Financial resilience: P&C business runs on millions of transactions, yet many firms still manage them on spreadsheets. That lengthen settlement times and credit risk. Automating financial operations significantly strengthens resilience.
- Revenue and partnerships: Automating data transfer between insurers, distributors and partners improves both the partnership experience and the end customer experience.
Immediacy is a business decision, not just a tech one
Insurers are getting closer to real-time financial and operational decision-making, but they aren’t there yet. Three things need to happen first: data needs to be brought into shape and democratised so teams can access it quickly, that data needs to run through the full process rather than isolated decision points, and firms need to understand where real-time actually serves the customer.
That last point matters. Some customers want their claims paid instantly. For others, a claim settled in seconds can feel like it was never properly reviewed. Real-time capability is powerful but knowing when to use it is a business judgement.
Governance and security build the trust to move faster
Strong data governance and ownership are fundamental. If you don’t know where your data sits, who owns it or how it’s managed, it’s very hard to trust it. Insurers with complete governance controls and full auditability across the data lifecycle operate with a far stronger control framework.
Security plays the same role. When teams trust their data, they share and act on it more confidently, including when they need to exchange it securely with external partners.
From copilot to agentic, with humans in the loop
AI has moved quickly from assisting employees to agentic capability, where agents can take autonomous actions to drive defined outcomes. But the human stays at the centre.
Some decisions will need full human supervision. Others will see agents surface recommendations that people validate against the source. And some well-defined, repeatable tasks can run end-to-end autonomously. The balance is a business risk decision: how much to hand to autonomous engines, and where human expertise must remain.
For firms just getting started, the priorities are clear:
Get the data right first. It needs to be in shape and easily accessible, usually through a platform that connects different data sets efficiently.
Know where data creates value. Understand which processes and services it can improve.
Decide your ambition. Some insurers embed AI into existing processes for quick wins through pilots and POCs. Others are becoming frontier organisations, redesigning around an AI-first North Star and working back from there.
And crucially, this is as much a people function as a technology one. Insurance is naturally risk-averse, with deep expertise built over decades. Change management, upskilling and helping people see how they benefit are just as important as the tech itself.
The power of the partnership
The AutoRek and Microsoft partnership comes together around two things: trust and experience. Microsoft brings scalability and security at the platform level. AutoRek brings highly compressed data reconciliation, native integration and full auditability.
What makes it work is a shared focus on customer outcomes. Rather than three parties each pursuing their own goals, customers, AutoRek and Microsoft co-create around a single question: what is the biggest pain point, and how will success be measured? When everyone drives towards the same outcome, the results speak for themselves.
Watch the full interview here: