- Key Insight: The real reason agencies struggle with AI is disconnected workflows.
- Supporting Data: 10 to 30 minutes added to routine policy transactions.
AI is supposed to make tasks easier, faster and more efficient. But an improper AI deployment can miss the most obvious time-savings.
Disconnected workflows, such as manual document reviews or re-entry of information across carriers, can add 10 to 30 minutes to routine transactions, according to an
Fleets are also adopting AI for uses such as driver risk analysis and video interpretation, but they are not sharing the information as quickly with brokers as they do with insurers, according to
Read more in Digital Insurance's recent coverage:
Disconnected systems add 30 minutes to routine policy tasks
Workflow integration — not AI adoption — is the central obstacle for insurance agencies, according to an Augie Group survey of 600-plus insurance professionals. While 79% of agencies use AI, 60% cite limited system integration as their top operational challenge, with disconnected workflows adding 10 to 30 minutes to routine policy transactions. Half of respondents call quoting the most manual process, followed by submissions (39%), renewals (36%) and endorsements (29%). Agencies can accelerate progress by documenting and quantifying time lost to redundant data entry — then raising those metrics directly with carrier field representatives to build the business case for integration improvements.
Read more:
60% of fleets use AI, but deep integration remains rare
Despite 60% of fleets currently using AI, most deployments remain surface-level, according to SambaSafety's 2026 Telematics Report, which surveyed 740 fleet, broker and carrier representatives. The most common applications are driver risk analysis (34%), video interpretation (31%) and administrative automation (24%). Adoption is expected to grow — 49% of fleets anticipate using AI for driver risk analysis in the future. A key gap: fleets are sharing telematics data with insurers faster than with brokers, even though brokers are often better positioned to act on it. Building client trust around data use — explaining the why, how and value — is the critical first step to closing that disconnect.
Read more:
AI in core systems, not bolt-ons, is the real modernization play
ForceAI founder Adam Ibrahim warns that bolt-on AI tools and demo-driven insurtechs are eroding carrier trust industry-wide, making it harder for legitimate solutions to gain traction. His advice: build AI directly into core systems, not layered on top of them. Ibrahim sees AI-driven efficiency gains coming incrementally — single-digit improvements over time — not in the sweeping transformation many predict. The self-funded, Florida-based startup, founded in 2024, integrates its unified P&C and health platform in months rather than years and keeps carrier data in isolated environments, avoiding model training on client data.
Read more:
Data modernization lets underwriters shift focus to risk
Siloed data and legacy systems remain the primary barriers preventing insurers from scaling generative AI applications, according to AWS financial services lead John Kain. The path forward requires centralizing governance and compliance infrastructure while distributing innovation capabilities across the organization. On the underwriting side, AI-enabled data processing is eliminating a significant bottleneck: underwriters reading hundreds of pages of engineering reports before a submission can even begin. Technology that handles unstructured data analysis allows underwriters to concentrate on portfolio-level risk decisions in real time — shifting their role from data analyst to risk strategist. Customer experience expectations, increasingly shaped by retail interactions, are accelerating this transformation.
Read more:
AV insurers face gaps in AI decision data, policy terms
Autonomous vehicle AI operates through three distinct model layers — perception, prediction and planning — each with its own failure modes that complicate underwriting and claims. Without access to AV decision logs, pricing risk becomes significantly harder than with conventional telematics data. When accidents occur, insurers must determine whether losses fall under auto, product liability or cybersecurity coverage, particularly when over-the-air communications are involved. Existing policy language may not be fit for purpose, requiring revision to address how AV developers program risk-weighted decisions. Tracking NAIC's Model AI Bulletin guidance and its AI evaluation initiative is critical as regulatory frameworks continue to evolve.
Read more:
MGAs should tie AI ROI to speed to revenue, not cost savings
With 60% of organizations planning to increase AI investment over the next year — yet one in five currently limiting use due to operating costs — MGAs need a more precise framework for measuring return. Rather than counting deployed capabilities or tracking hours saved, the stronger metric is speed to revenue: how quickly AI helps a team move from identifying an opportunity to writing business. Practically, that means targeting AI at submission prioritization, reducing manual underwriting handoffs and compressing the workflow build-out required to enter new markets. Preserving institutional underwriting knowledge as experienced staff retire is an additional long-term consideration worth factoring into ROI calculations.
Read more:
AI shopping tools may erode the loyalty buffer, JD Power finds
With 66% of homeowners saying switching insurers is too difficult or time-consuming, carriers currently benefit from inertia — but JD Power warns AI-powered shopping tools could quickly erode that advantage. Trust, seamless service and channel effectiveness are the strongest retention levers, according to the 2026 U.S. Home Insurance Study of 17,191 respondents. Proactive outreach measurably improves policy comprehension: customers contacted in the past year understood their coverage at a 60% rate, versus 48% for those with no insurer contact. Only 55% of homeowners fully understand their policy — a figure essentially unchanged over six years. Amica ranked first in homeowners satisfaction; State Farm led in renters.
Read more:
Insurers weigh blocking Meta's Muse as scraping concerns grow
Insurance marketplaces are drawing lines around Meta's Muse AI agent after Insurify's co-CEO found the tool bypassing bot detection, concealing its identity and stripping carrier details from quotes — behaviors he compared to hacking. Insurify has blocked Muse and is negotiating terms with Meta; Kin, another marketplace has opted to allow access. For carriers and marketplaces evaluating Muse, Insurify's experience points to a concrete standard: require API or model context protocol access, enforce terms-of-service compliance and demand bot self-identification. Without those guardrails, incomplete quotes risk misleading consumers on coverage details and distorting quote-to-bind ratios for carriers.
Read more:
This roundup was created with AI assistance and human review. Introductory bullet points created by AI with editorial review.
