Insurance executives report concerns about bias in artificial intelligence systems, yet most expect claims administration to be handled end-to-end by AI within the next two years.
Research from embedded insurance provider EIP found that 87% of senior insurance professionals across the UK and Europe are concerned about bias or unfair outcomes in AI-driven processes. At the same time, 90% of respondents expect claims administration to be managed entirely by AI within the next 24 months.
The findings come as insurers expand the use of automation across underwriting, claims handling and customer service. AI systems are already used to analyse documents, triage claims and support underwriting assessments, allowing insurers to process large volumes of information more quickly while retaining human review of final decisions.
The EIP study surveyed 250 senior insurance professionals across the UK and Europe.
Despite expectations of wider automation, insurers remain cautious about applying AI to certain functions.
Claims submissions were identified as the area where the industry feels least comfortable introducing automation, cited by 40% of respondents. Underwriting recommendations followed at 39%, while 35% reported reservations about automating customer interactions.
These concerns align with wider industry discussions about the role of automated decision-making in claims and underwriting. Analysts have warned that biased data or flawed models could expose insurers to disputes if automated systems influence claims outcomes.
The survey identified bias as the primary concern linked to AI deployment.
Other risks cited by respondents included data security and privacy (23%), regulatory non-compliance (21%), and system reliability or errors (21%). Job displacement or staff resistance was mentioned by 20%. Among larger insurers, regulatory concerns ranked higher, with 33% identifying compliance issues.
Regulators are increasingly examining the use of AI across financial services. In the UK, policymakers have proposed a context-based approach to regulating AI technologies. In Europe, the EU AI Act will introduce a risk-based framework for artificial intelligence systems when it takes full effect in 2026, including obligations related to transparency and oversight.
Among survey participants, 39% said transparent algorithms and decision logs could help address concerns about AI use in insurance.
Industry observers say explainability and governance are becoming central requirements as insurers adopt AI tools. Systems that produce automated recommendations must allow insurers to understand how decisions were reached, particularly where claims or underwriting outcomes may face scrutiny.
Human involvement remains widely expected in AI-supported decision-making. According to the survey, 99% of respondents said some level of human oversight should accompany AI-driven outcomes.
Despite reservations about certain applications, insurers expect AI to influence the sector.
All respondents in the survey said AI will have a sizeable impact on the insurance industry. Of those, 58% expect the technology to deliver efficiency gains, while 33% believe it will transform the industry entirely.
Insurance firms are exploring AI across multiple business functions. Customer service and chatbots were cited by 45% of respondents already using or evaluating AI in this area. Risk modelling and data analytics followed at 43%, while 41% reported activity in claims management. Underwriting and pricing and fraud detection were each cited by 40%, while 38% reported using or exploring AI for marketing.
When assessing AI tools, insurers identified ease of integration as a leading consideration. Around 27% said simple integration is the most important factor, while 27% cited partnership and support. Only 10% said cost and return on investment would strongly influence their decision.
“Insurance is, by necessity, a heavily regulated industry, which makes introducing AI complicated. This research shows that the development of industry-specific capabilities could help to overcome challenges and accelerate the adoption of AI right across the industry, whether that’s for customer service, risk modelling or even claims management and decisioning,” said Ross Sinclair, founder and CEO of EIP.
“It is important to remember that bias is an issue for insurers regardless of whether they are using AI or not. It has always been a challenge for the industry,” said David Mitchell-Dawson, director of product at EIP. “Insurers told us that they’re expecting their full end-to-end claims administration to be handled by AI within the next 15 months on average, so frankly if you’re not doing anything now, you’re already too late.”
The survey was conducted by Censuswide on behalf of EIP between January 28 and February 4, 2026, involving 250 senior insurance professionals across the UK and European markets including Italy, Germany and Spain.
