Scaled artificial intelligence (AI) is starting to reshape property and casualty (P&C) claims, but most carriers remain in early-stage experimentation, according to Bain & Company’s 2025 Claims Maturity Assessment.
The global study suggests many insurers have yet to move beyond pilots into broad operational deployment.
The assessment, conducted with GLG and Dialectica, surveyed 81 P&C insurers worldwide. It found that 78% of respondents are using generative AI in some capacity, but only 4% have scaled the technology across their organisations.
Current deployments are mostly narrow and task-specific, aimed at easing manual work in core claims functions. Common use cases include summarising lengthy documents, supporting customer communications and helping to detect potential fraud.
Read more: Insurers expand AI use
Bain’s findings track with a separate Morningstar DBRS report showing that insurers are broadening AI use in areas such as underwriting and back-office operations to improve efficiency and reduce costs, even as they face growing scrutiny around governance and customer impact.
In that research, carriers are turning to AI to automate tasks like generating policy templates, summarising interactions and analysing documents, which parallels the way generative tools are being applied to reduce manual effort in claims.
Bain’s analysis contends that the larger opportunity lies in redesigning the claims journey around generative AI, rather than sprinkling tools into existing workflows. At scale, the firm estimates that AI can substantially shorten settlement cycles across a wider range of claims, cut home claims processing times in half and lift productivity by 35%.
Despite that potential, only 27% of insurers in the survey said they are pursuing what Bain describes as a full claims transformation. Many organisations reported constraints tied to limited in-house AI expertise, concerns over model accuracy and ongoing questions around data security and privacy.
Read more: Insurance’s gen AI reckoning has come
The report highlights five recurring characteristics among insurers that are moving beyond pilot mode. One is setting a clear ambition to rethink the claims process end to end, instead of adding AI on top of legacy systems and manual steps.
Another common factor is maintaining a roadmap for digital tools, including when to build capabilities internally and when to buy from external vendors. These carriers also emphasise close coordination between claims and technology teams to reduce handoffs that can slow delivery.
Broader market research from Deloitte indicates many insurance and financial services executives now expect AI initiatives to take two to four years to deliver satisfactory returns, a longer payback period than typical IT projects and a possible reason some insurers hesitate to commit to full-scale claims overhauls.
Across carriers and brokers, the strongest early gains are emerging in efficiency-focused applications such as faster quote generation, reduced claims cycle times and higher employee productivity, which aligns with Bain’s view that near-term value is concentrated in operational use cases.
Bain notes that leading organisations also encourage broader participation in innovation from claims staff, rather than concentrating generative AI responsibility solely within IT or data science units. They invest in workforce planning as well, with a focus on future skills in data, analytics and digital enablement within claims operations.
While most P&C insurers are still “dabbling” in generative AI, the report says those pushing toward scaled, integrated deployments are beginning to see measurable benefits. According to Bain, these early movers are reporting gains in speed, unit cost and productivity across multiple parts of the claims lifecycle.
