Sponsored by Xceedance
Three Critical Factors Insurance Companies Must Address Before Integrating AI

Artificial intelligence is rapidly transforming the insurance industry, from underwriting and claims processing to customer engagement and risk modeling. Yet for all the promise AI holds, the path to successful implementation is far from straightforward. Insurers that rush to adopt new technology without a deliberate strategy often find themselves grappling with internal resistance, unreliable outputs, and compliance headaches.
The pressure to modernize is real. Insurers across the market are investing in AI-driven tools to gain efficiency, improve accuracy, and stay competitive. But the companies seeing the strongest returns are those that treat AI adoption not as a technology project alone, but as a broader organizational transformation.
“AI can be an incredibly powerful tool for insurance companies, but only if the foundation is right,” said Sachin Kulkarni, Executive Vice President, Commercial & Specialty Insurance and MGAs, Americas, Xceedance.
“You can have the most sophisticated algorithm in the world, and it won’t deliver value if your people aren’t prepared, your data isn’t clean, or your compliance framework hasn’t kept pace.”
Kulkarni, whose company partners with insurers to support technology-driven transformation, pointed to three factors that every carrier should evaluate carefully before bringing AI into core operations.
Change Management, Data Integrity, and Regulatory Oversight

Sachin Kulkarni, Executive Vice President, Commercial & Specialty Insurance and MGAs, Americas, Xceedance
1. Change Management
Perhaps the most underestimated challenge in any AI rollout is the human element. Insurance is a relationship-driven business, and many professionals who have spent decades honing their expertise may view AI with skepticism—or outright resistance.
“If you don’t bring your people along on the journey, the technology will sit on a shelf,” Kulkarni said. “Change management is not just an HR initiative. It has to be woven into every phase of the implementation, from initial planning through post-launch optimization.”
Effective change management starts with clear communication about why AI is being introduced, what it will and will not replace, and how it will support—not supplant—the work employees already do. Underwriters, claims adjusters, and actuaries need to understand how AI tools fit within their existing workflows and where human judgment remains essential.
Training is equally critical. Deploying a new system without investing in hands-on education and ongoing support creates confusion and erodes trust. When employees feel empowered to use AI as a resource rather than threatened by it as a replacement, adoption rates climb significantly.
“The companies that succeed are the ones that frame AI as an enabler,” Kulkarni said. “When your team sees how it removes tedious tasks and lets them focus on higher-value work, the resistance starts to fade.”
2. Data Integrity
AI is only as good as the data it consumes. For insurance companies—many of which operate on legacy systems that have accumulated decades of inconsistent, siloed, or incomplete records—data quality is a significant hurdle.
Models trained on flawed data will produce flawed outputs. In an industry where pricing accuracy, loss projections, and regulatory filings depend on reliable information, the consequences of poor data integrity can be severe. Inaccurate risk assessments, mispriced policies, and faulty claims decisions all trace back to the same root cause: garbage in, garbage out.
“Before you even think about what AI can do for you, you need to take a hard, honest look at your data environment,” Kulkarni said. “That means auditing your data sources, standardizing formats, eliminating duplicates, and establishing governance protocols that keep quality high over time.”
Data governance is not a one-time exercise. As new information flows into the system—from third-party sources, IoT devices, customer interactions, and more—insurers need ongoing processes to validate, clean, and monitor that information. Without those safeguards, even the most advanced AI tools will degrade in performance.
Kulkarni emphasized that data readiness is also about accessibility. Information locked in departmental silos prevents AI models from drawing the comprehensive insights that make them valuable. Breaking down those barriers, while maintaining appropriate security and access controls, is a prerequisite for meaningful AI deployment.
3. Regulatory Oversight
Insurance is one of the most heavily regulated industries in the United States, and the introduction of AI adds new layers of compliance complexity. State regulators are paying close attention to how insurers use algorithmic decision-making, particularly around issues of fairness, transparency, and consumer protection.
Several states have already introduced or are considering guidelines around the use of AI in underwriting and rating. The National Association of Insurance Commissioners has also been actively developing frameworks to address AI governance. Insurers that implement AI without accounting for this evolving regulatory landscape risk running afoul of requirements they did not anticipate.
“Regulatory scrutiny around AI is only going to increase,” Kulkarni said. “Insurers need to build explainability into their models from day one. If you can’t clearly articulate to a regulator how your AI reached a particular decision, you have a problem.”
Transparency is the operative word. Insurers must be able to demonstrate that their AI tools do not introduce or amplify bias, that decisions can be audited and explained, and that consumers are treated fairly regardless of the technology behind the scenes. This requires not only technical capabilities but also cross-functional collaboration between data science teams, compliance officers, and legal counsel.
Documentation is another essential component. Every model, every data input, and every decision pathway should be recorded and retrievable. Building this discipline into the AI development process from the start is far easier than retrofitting it after deployment.
How Xceedance Supports Insurance Partners Through AI Implementation
Recognizing that these challenges are interconnected, Xceedance works alongside insurance company partners to develop implementation strategies that address all three factors in a coordinated manner.
Rather than offering a one-size-fits-all technology solution, Xceedance takes a consultative approach, beginning with a thorough assessment of each partner’s organizational readiness, data environment, and compliance posture. This diagnostic phase helps identify gaps and priorities before introducing any AI tool.
“We don’t believe in deploying technology for technology’s sake,” Kulkarni said. “We sit down with our partners, understand their specific challenges and goals, and build a roadmap that accounts for their people, their data, and their regulatory obligations.”
On the change management front, Xceedance helps insurers develop communication plans, training programs, and feedback loops that keep employees engaged throughout the process. The goal is to build internal champions who understand the technology and can advocate for its value across the organization.
To ensure data integrity, Xceedance supports partners in auditing existing data assets, implementing governance frameworks, and establishing ongoing monitoring processes to maintain quality. This groundwork ensures that, when AI models are deployed, they operate on a reliable foundation.
On the regulatory side, Xceedance stays current with the evolving compliance landscape and helps partners build the documentation, explainability, and audit capabilities that regulators increasingly expect. This proactive approach reduces the risk of costly surprises down the road.
“Our role is to be a long-term partner, not just a vendor,” Kulkarni said. “AI implementation is not a one-and-done project. It requires continuous refinement, and we stay engaged with our partners well beyond the initial deployment to make sure the technology is delivering the results they need.”
Looking Ahead
As AI capabilities continue to advance and regulatory frameworks mature, the insurance companies that invest in thoughtful, well-supported implementation today will be best positioned to capture long-term value. The insurers that treat AI adoption as a strategic initiative—one that encompasses people, processes, data, and compliance—will separate themselves from those that chase the latest technology trend.
“The opportunity is enormous, but so is the responsibility,” Kulkarni said. “Insurance companies that get this right will be more efficient, more accurate, and better equipped to serve their policyholders. The key is doing it the right way from the start.”
To learn more, visit www.xceedance.com.
![]()
This article was produced by the R&I Brand Studio, a unit of the advertising department of Risk & Insurance, in collaboration with Xceedance. The editorial staff of Risk & Insurance had no role in its preparation.

