Welcome to this week’s issue of Risk, Bytes and Beyond - your bi-weekly roundup of the latest insights at the intersection of data analytics, artificial intelligence and insurance. Delivered to your inbox every second Friday! In each edition, you’ll find:
Curated content from experts in data and insurance
An insight brief on a changing topic
The latest headlines on data and AI in insurance
Thanks for reading - let’s dive in!
💡 Noteworthy Content
Short one-hour webinar on predicting medical costs with linear regression in Python, using a dataset from Kaggle. Great for beginners, to brush up on fundamentals, or for business folks interested in insurance-related machine learning.
Great article from BCG outlining the tech foundations insurers must build to succeed in embedded insurance. It emphasizes that to tap into the $87 billion (and growing) embedded-insurance market, carriers need a modular, API-driven stack that supports real-time data, dynamic underwriting, policy issuance, and claims processing.
💭 Insight Brief
Legacy tech isn't just an IT problem - it's a data problem.
Legacy tech - defined as outdated technology that does not meet modern user requirements - is a major roadblock for innovation in insurance.
Many insurers struggle with legacy tech, such as mainframe systems, applications developed in COBOL, monolithic software without modern API interface and siloed data storage.
Legacy tech in insurance causes many problems, such as high maintenance costs, limited flexibility and scalability of IT systems, cyber threats and compliance issues.
There have been recent estimates by PwC, that insurers spend up to 70% of their IT budget on maintaining old legacy technology. Money that is then lacking for modernization and innovation.
One often overlooked problem of legacy tech relates to data analytics and artificial intelligence projects. Legacy tech leads to problems in every step of the typical lifecycle of a data product:
Access and Ingestion: Monolithic legacy systems tend to produce siloed and fragmented data, making it difficult to locate, integrate, and clean. As a result, data engineers, analysts, and scientists spend significant time hunting through various databases, warehouse tables, and operational systems just to gather the data they need.
Monitoring and Cleaning: Fragmented data is hard to manage and govern. It becomes difficult to monitor quality, assign clear ownership, and establish consistent KPIs and metrics. Without reliable data governance, trust erodes - and without trust, data-driven projects stall or fail entirely.
Analysis and Experimentation: Even when data teams overcome access and quality issues, they might need to work with outdated tools themselves. Legacy environments limit the ability to analyze data effectively, experiment freely, or build modern machine learning models.
Deployment: Finally, once a data product is ready, it may be nearly impossible to deploy within a legacy tech stack. These systems often lack the flexibility and integration capabilities needed to support modern data products, such as predictive machine learning models.
In short, legacy technology doesn't just slow down IT operations - it actively undermines insurers' ability to innovate, compete, and make use of their most valuable asset: data.
📰 Industry Headlines
Axa and Allianz lead the new AI Index for Insurance report, which ranks 30 insurers across North America and Europe. According to the report, insurers are investing in AI mainly in the following use cases: claims automation, customer service, underwriting, and internal knowledge search.
A new report finds insurers are the least advanced in AI, with many stuck at beginner levels and held back by data protection, security and talent hurdles. As a result, they’re failing to fully harness AI for risk assessment and claims evaluation - leaving valuable profit on the table.
Towergate Insurance warns that rising AI use may lead to hallucinations that spread misinformation - or even outright fraud - stressing the need for media verification and internal fact‑checking by professionals. They also flag data exposure risks when business info enters public AI tools, urging companies to bolster internal safeguards and cyber‑insurance policies.
Was this email forwarded to you? Sign up here!