Page 71 - IMDR Journal 2025

Research Article
mistrust and dissatisfaction, particularly when claims are explanations for decisions.
rejected without clear explanations.
Bias in Claim Decisions
Perceived Efficiency of AI-Driven Claims Processing Another major challenge is the perception of bias in AI-
A majority of respondents (47.5%) perceive AI-driven driven claims processing. 29.4% of respondents reported
claims processing as much better than traditional methods, concerns about bias, particularly in claim rejections. For
while 45.1% rate it as slightly better. This indicates that example, older respondents and those from rural areas felt
92.6% of respondents find AI-driven systems more efficient that their claims were disproportionately rejected. One rural
than traditional methods. respondent stated, "I feel like the system is biased against
people like me who live in villages. My claim was rejected,
● Much Better 25.5%
but my friend in the city had a similar claim approved." This
● Slightly Better 45.1%
suggests that AI algorithms may inherit biases from the data
● Neutral 25.5% they are trained on, leading to unfair outcomes for certain
● Slightly Worse 0% demographic groups.
● Much Worse 0% Technical Glitches
Respondents appreciated the faster processing times, with 52.9% of respondents reported experiencing technical
51% of respondents reporting being very satisfied and glitches with AI-driven systems, such as chatbots providing
29.4% being satisfied with the time taken to process claims. incorrect information or AI systems failing to process claims
For example, one respondent stated, "My claim was correctly. These glitches not only reduce efficiency but also
processed in just 2 days, which was much faster than the 10 erode consumer trust. For example, one respondent noted,
days it took last time without AI." "The chatbot gave me the wrong information, and I had to
call customer service to fix it. It was frustrating." Addressing
However, 9.8% of respondents were dissatisfied, and 9.8%
these technical issues is critical to improving the reliability
were very dissatisfied, indicating that while AI is generally and user experience of AI-driven systems.
perceived as efficient, there is room for improvement,
particularly in addressing technical glitches and improving
user experience.
Perceived Fairness and Trust in AI-Driven Claims
Trust in AI-driven decisions is mixed, with 45.1% of
respondents expressing trust, 27.5% remaining neutral, and
27.5% distrusting AI systems. Key factors influencing trust
include
● Transparency: 29.4% of respondents cited a lack of
transparency as a major challenge.
Fig 5.2.1
● Bias: 29.4% of respondents reported concerns about bias
in AI decision-making, particularly in claim rejections. Impersonal Interactions
For instance, one respondent shared, "My claim was Finally, 17.6% of respondents cited impersonal interactions
rejected without any explanation. It felt like the system was as a challenge. While AI systems are efficient, they often
biased against me." This highlights the need for insurers to lack the empathy and human touch that customers value,
adopt explainable AI (XAI) techniques to provide clear and particularly in sensitive situations like health insurance
claims. One respondent stated, "The chatbot was quick, but
understandable explanations for AI-driven decisions.
it felt cold and impersonal. I would have preferred talking to
Fairness was another area of concern, with 45.1% of a human." This suggests that while AI can handle routine
respondents feeling that AI systems were fair, while 27.5% tasks, human oversight is still necessary for complex or
reported concerns about bias. Older respondents and those emotionally charged cases.
from rural areas were more likely to perceive AI systems as
unfair, suggesting that demographic factors play a role in Comparative Analysis
shaping perceptions of fairness. AI vs. Human-Driven Claims Processing
Challenges Identified The survey findings reveal significant differences between
Lack of Transparency AI-driven and human-driven claims processing in terms of
efficiency, accuracy, and customer satisfaction.
The most significant challenge identified by respondents is
● Efficiency AI-driven systems are much faster, with claims
the lack of transparency in AI-driven claims processing.
processed in 2-3 days compared to 10-15 days for traditional
54.9% of respondents reported experiencing claim
rejections without clear explanations, leading to frustration methods. This speed is a major advantage, particularly in a
and mistrust. One respondent noted, "I have no idea why my high-volume market like India.
claim was rejected. The system just said 'rejected' without ● Accuracy AI systems are more accurate, with a 90%
any details." This lack of transparency erodes consumer accuracy rate compared to 75% for manual processing. This
trust and highlights the need for insurers to adopt reduces the likelihood of errors and improves customer
explainable AI (XAI) models that provide clear, jargon-free satisfaction.

62