83% of insurers would let AI run repeatable operational tasks, ISG study finds

# What the study measured The ISG study surveyed senior leaders across underwriting, operations, claims, technology and transformation in North America, Europe and Asia. It covered 20 operational activities across submission intake and triage, quote generation, bordereaux processing, claims adjudication and compliance screening.

# What insurers say they want

# What 'governed' means in practice The study draws a specific distinction between general-purpose AI and a 'governed' model. A governed model:

• uses controlled policy wording, endorsements, underwriting appetite and claims guidance,

• allows a named person to override or stop actions,
• operates within the institution's appetite, thresholds and referral triggers.

This governance framing helps explain why 86% of respondents said consequential decisions — those that create competitive difference — should remain with people, even when AI handles the routine work underneath.

# Why insurers see capacity benefits Carriers estimate one in nine broker submissions are declined or left unquoted not because of appetite but because operations cannot process them in time. Where AI is already in operations, respondents reported measurable operational benefits: 61% saw productivity improvements and 51% reported faster cycle times. Respondents expect operating costs to fall about 16% over two years.

Martin Henley, CEO of mea Platform, summarised the conditional appetite: insurers are ready to hand repeated work to AI if the model understands insurance and runs inside their rules. He added that capacity unlocked by AI matters most for what insurers choose to do with it, such as expanding appetite or entering new lines.

# The adoption gap: ambition vs reality Nearly all insurers (96%) have AI-led operational redesign on their agenda. Yet the portion that defines an advanced operating posture — where AI is central to workflow design — stands at just 13% today. Fully AI-native operations, in which AI executes defined processes end to end while people set policy and manage exceptions, is below 1% today and only 12% expect to reach that within two years.

# Practical implications for insurers and brokers

# Bottom line