In 2025, 27% of Surety carriers weren’t using AI at all. In 2026, no carrier surveyed reported zero AI use. It’s one of the clearest shifts in Surety Outlook 2026: Tech & Ops Benchmarks, Everest Group’s survey of North American Surety carriers and brokers.
The adoption question is settled. The value question isn’t. Only 15% of carriers use AI widely across their workflows and decision-making. For everyone else, it’s what Vinod Kachroo, Head of Americas Business at Tinubu, calls “bolt-on AI”: AI that handles an isolated task, then hands its output back to the same emails, re-keying and handoffs it was meant to fix.
AI Is Speeding Up Steps, Not Processes
Carriers’ top AI priorities are financial statement analysis and rating and pricing (62% each). Brokers are pointing AI at their most manual work, led by data validation and enrichment (38%), then placement routing, portal entry and reporting (31% each).
That task-level AI creates real value. Carriers are already using it to read financial statements and import bonds issued outside their systems. But speeding up one step does little if the rest of the process stays the same. Ingestion should be the starting point, not the finish line.
What’s Holding AI Back
The problem isn’t the AI. It’s what it sits on: older systems, patchy data and fragmented communication between carriers and brokers. “AI has to operate within a context,” says Kachroo. “It is difficult for carriers to move AI use beyond the task level if the rest of their system and workflow are not modern enough to facilitate it.”
The survey shows where the gap sits. 92% of carriers report digitally integrated post-bind servicing, but only 31% report integrated quote exchange. So carriers are aiming AI at underwriting while the information feeding it still arrives by email and phone, and gets re-keyed along the way. AI can read a financial statement in seconds. It can’t speed up a submission stuck in someone’s inbox.
From Bolt-On to Built-In
Once data is digital and can move through a system, AI can move from reading documents to completing transactions. That’s where agentic AI comes in: AI that doesn’t just analyze information but carries out routine work end to end. “Once you have digitized your workflow, you can put agentic AI on those processes and get further value and efficiency,” Kachroo says. “It’s a virtuous circle.”
The practical place to start is routine, repeat business, where agentic AI can handle standard work and flag exceptions for an underwriter.
What Everest Group Recommends
For carriers: segment the operating model by complexity. Use standardized, lower-touch processing for routine business, and keep specialist underwriters for complex risks and referrals.
- What to track: turnaround for routine commercial bonds under 24 hours, and fewer delays on complex bonds from manual indemnity review (cited by 54% of carriers), internal capacity sign-off (46%) and reinsurance referrals (31%).
For brokers: use AI to reduce administrative effort. Start with data validation, placement support and portal entry, where manual work remains high.
- What to track: reductions in manual handling time across data validation, placement routing, portal entry and reporting.
The payoff is real. Carriers estimate automation could take on about a fifth of their day-to-day operating costs, roughly 7.5% of everything they spend. AI built into the workflow is part of how they’ll capture it.
AI adoption is no longer the question. The carriers that see results will be the ones that connect the work around it.
How does your AI adoption compare with your peers’? The full Surety Outlook 2026: Tech & Ops Benchmarks report breaks down where carriers and brokers stand on AI, connectivity and operating costs, with Everest Group’s recommendations for each.
