Overcoming the Top 3 Barriers to AI in Insurance

AI in insurance is advancing rapidly, but meaningful adoption still faces roadblocks. The truth is, successful AI isn’t just about the algorithms. It’s about people, process, purposeful design, and a human-first approach. 

In this post, we will break down the top 3 barriers insurers must overcome to successfully implement, scale, and unlock value from AI, and how to tackle a fourth emerging challenge: measuring ROI. 

Inconsistent or Low-Quality Data: Garbage In, Garbage Out 

AI is only as powerful as the data that fuels it.  

Disjointed systems, inconsistent terminology, and outdated processes lead to poor data quality, hindering AI performance and insight reliability. For AI to truly drive change, data must be clean, connected, and cloud-ready.  

This means that a successful AI journey depends on digital maturity: breaking down silos, improving data governance, and investing in infrastructure. Only then can AI realise its full potential. 

Cultural Resistance: “It’s Just Tech” Thinking 

AI isn’t here to replace people, it’s here to amplify human expertise. 

The real value of AI in 2025 lies in its ability to automate the routine, giving expert teams more time to focus on strategic decisions. That means: 

• People remain in control of when and how AI is applied. 

• Trust builds when teams co-design AI use cases 

Change is communicated clearly and transparently. When AI feels intuitive, useful, and human-centred, adoption follows naturally.  

Transparency and Trust: Bias, Privacy, and Hallucinations 

Trust isn’t optional, it’s essential. 

According to a recent study,75% of C-suite executives are concerned about data privacy and security breaches in the context of AI. These concerns highlight the need for a responsible AI framework that addresses risk from the ground up: 

• Bias and toxicity: Outputs are as biased as the underlying data they are trained on.  

• Data leakage: Employees entering sensitive information into OpenAI models risk exposing it publicly 

• Hallucination: AI generating convincing but factually incorrect responses 

To tackle this, AI must be explainable, fair, and auditable. Human oversight plays a critical role in: 

• Validating outputs 

• Spotting unintended bias 

• Ensuring that decisions remain ethical and compliant 

At inari, we believe experts shouldn’t just approve AI results, they should shape and challenge them to ensure real-world accuracy and fairness. 

(Bonus Barrier) Measuring ROI: How Do We Know It’s Working? 

72% of Data analytics, & IT leaders reported positive ROI from GenAI projects, yet only 59% say those results are measured quantitatively. 

Why? 

Because it’s hard to isolate AI’s impact from other tech efforts, and most insurers lack a clear success benchmark. 

To overcome this: 

• Define success metrics upfront 

• Align AI use cases to specific KPIs 

• Track both qualitative and quantitative outcomes 

Without clear metrics, AI’s true value risks being overlooked. 

AI at inari: Focused, Responsible, Human-Centred 

At inari, we take a practical, outcome-driven approach to AI. We don’t use it for the sake of it, we use it to solve real problems. 

Our cloud-based API-native Underwriting Workbench, Kitsune, offers flexibility and scale, empowering insurers to streamline workflows and improve underwriting accuracy. 

Where AI delivers real value, we apply it, such as: 

• Submission intake and triage 

• AI-generated insured descriptions 

• Conversational chatbot assistants 

And where other technologies are better suited, more accurate, or cost-effective, we use those instead. 

How inari Helps Break the Barriers 

• Accurate data capture: Start at the source for cleaner, richer insights, so that risk is properly assessed 

• Human-first UX: Designed so people can easily interact with and validate AI output in an efficient and effective way 

• Tech with purpose: Focused on speeding up submission-to-bind and supporting smarter decisions 

• Data privacy built-in: Self-hosted models keep everything in-house and secure; no client data used in training. 

• Compliance-first: We stay ahead of evolving AI regulations to protect users and customers 

Conclusion: The Future of Insurance is Human-Led AI 

AI works best when it’s designed with people, for people. 

When co-created and deployed responsibly, with clarity, accountability, and transparency, AI becomes more than a tool. It becomes a trusted partner. 

At inari, we’re building that future, where smart, ethical, and user-friendly AI drives real value for insurers and customers alike.