AI Executives Escaping the Limitations of Trial Runs
In the rapidly evolving world of technology, Artificial Intelligence (AI) is making significant strides in various industries, and the insurance sector is no exception. AI is revolutionising the insurance underwriting process, offering a more efficient and effective approach to risk assessment and customer experience.
Anand "Andy" Logani, the executive vice president and chief digital and AI officer at EXL, has been at the forefront of this transformation. He noted, "AI's value comes from being part of business workflows, as demonstrated in the insurance underwriting industry."
However, integrating AI into enterprise workflows is not without its challenges. According to a survey by EXL, about 60% of AI initiatives across various industries are stuck in the pilot phase, with issues such as a lack of talent, incompatible legacy systems, and siloed data cited as obstacles.
To overcome these challenges, AI leaders are adopting strategic, phased approaches that emphasise practical business value. They are focusing on solving specific business problems rather than impressive tech demos. Successful AI implementation requires experts who understand the intricacies of the business and have senior-level support to reengineer workflows.
Detailed Challenges and Solutions
Data Quality and Availability
AI systems require clean, consistent, and integrated data. However, many enterprises have fragmented, incompatible, or incomplete data sources. To overcome this, AI leaders implement robust data integration and cleansing processes, start with projects addressing specific business problems, and build scalable data infrastructure.
Legacy Systems Integration
Existing software and IT infrastructure often cannot support modern AI models or APIs. To bridge this gap, AI leaders develop middleware or APIs to bridge old and new systems, plan incremental upgrades, and focus on integration that supports business workflows.
Skills and Talent Gaps
AI requires specialized expertise that many IT teams lack. AI specialists may not understand business context. To address this, AI leaders invest in cross-functional training, hire AI specialists with domain knowledge, and foster collaboration between IT and business teams.
Security and Governance
Concerns about data security, regulatory compliance, and AI decision-making governance slow adoption. To establish trust, AI leaders establish clear AI governance frameworks, integrate AI compliance into existing security protocols, and maintain transparency and explainability.
Organizational & Structural Misalignment
Traditional hierarchies conflict with AI-driven cross-functional collaboration. To resolve this, AI leaders design AI centres of excellence balancing autonomy and coordination, redefine workflows and accountability to support AI adoption, and apply change management practices.
Employee Resistance and Cultural Barriers
Fear of job displacement and lack of trust in AI causes resistance and hampers adoption. To build trust, AI leaders communicate transparently about AI goals and impacts, create psychological safety for experimentation, and invest in training and cultural transformation.
Budget and ROI Uncertainty
High upfront costs with unclear or delayed returns lead to skepticism and stalled projects. To address this, AI leaders use realistic financial planning including full cost of data prep, talent, and maintenance, start with MVPs that demonstrate value quickly, and develop new AI ROI metrics.
Additional Insights
Successful AI integration increasingly requires fundamental shifts in workflow design, moving towards autonomous agentic systems with embedded intelligence at the operational level. Despite wide interest, about 95% of generative AI pilots fail to produce rapid revenue growth, primarily because of these integration and organisational challenges. Companies that successfully navigate these challenges demonstrate a balance between immediate gains and building long-term AI capabilities, with strong leadership commitment and sustained change management.
In the insurance underwriting process with AI, the underwriter analyses the comprehensive AI output in the system and issues policies and manages customer engagements. By overcoming the challenges and adopting these strategies, businesses are reimagining their entire operating models to accommodate the use of AI, going beyond pilot projects.
[1] EXL Survey, 2023 [2] Forbes Technology Council, 2024 [3] S&P Global Research, May 2025 [4] McKinsey & Company, 2026 [5] Deloitte Insights, 2027
Read also:
- EA Relies on Madden and Battlefield to Drive Microtransactions Recovery
- Expense for Creating a Digital Platform for Fantasy Sports
- AI-Enhanced Battery-Swapping Station in Southeast Asia Officially Opens Its Doors
- Honda unveils blueprint for design, advanced driver assistance systems, electric vehicles, fuel efficiency, and technology development
