This month’s roundup covers the AI news, insights, and industry trends shaping what's next in insurance.
Latest Articles as of November 5
News: Consultancy says claims data can sharpen underwriting
The brief: Oxbow Partners, a management consultancy focused on the insurance sector, says in a new report that AI could go beyond claims efficiency and generate information for underwriting, pricing, portfolio performance, and risk evaluation. The report notes that claims records hold loss outcomes and the effects of underwriting decisions, but insurers have struggled to extract insight because of fragmented systems and unstructured documentation. It says AI is already used in intake, document interpretation, case prioritization, and handler support, and sees wider uses in predicting claim development, identifying recoveries, and spotting changes in loss experience. The partner who wrote the report said claims teams sit on some of the richest data in insurance, yet most of it has never been properly used. The report adds that priorities should depend on each insurer’s operating model and portfolio, and that data quality is a precondition.
News: Hiscox cuts broker-to-quote time 75% with AI
The brief: Hiscox says it has cut the time between broker submission and quote by about 75% in its terrorism and architects and engineers lines, and its AI underwriting tool is now expanding across US retail. Insurance Business reads the result as one that affects conversion, not just process. Hiscox USA written premium rose 8.4% to $791.6 million for the nine months to September 30, with direct and digital premium up 10.1% to $491.4 million and broker premium up 5.8% to $300.2 million. The company says its change program is on track for $75 million of profit and loss benefit this year and still expects 9% constant-currency growth for the full year. The article adds that the update suggests AI-driven discovery is beginning to affect how small business owners find and buy insurance.
News: Strategists expect states to lead on AI rules
The brief: Speaking at the APCIA annual meeting in Boston, two political strategists, a former Obama chief strategist and a former Bush deputy chief of staff, said data centers have become a recurring campaign issue ahead of the midterm elections. The former Bush deputy chief of staff expects state-by-state governance because federal oversight is limited, and he pointed to the Texas governor’s moratorium, which asks developers to show water and power plans, capacity sharing, local buy-in, and property-tax commitments without tax incentives. He said federal AI regulation is unlikely in the next two years, while the former Obama chief strategist said states will likely step in, though he thinks presidents should lead on setting rules. The former Obama chief strategist acknowledged that AI’s power demands require the infrastructure, but said residents need a sense of participation.
News: Americans want human review of AI claim decisions
The brief: A Pew Research Center survey of 3,488 US adults finds Americans are generally open to AI for lower-stakes tasks such as translation, and that they want to know when AI is used in their care. Only 5% would accept insurance claim decisions made by AI without human review, and almost half said AI should play no role in those decisions. Sixty percent said they are uncomfortable with AI helping a doctor make a diagnosis or treatment plan, and patients trust doctors to use AI in their interest more than insurers, government, tech firms, or hospital leadership. Pew’s director of science and society research, who led the report, said people generally don’t want AI operating solo and that transparency is key.
News: Underwriters ask for AI tool inventories and owners
The brief: Underwriters are asking more about how companies manage AI agents, and Lockton’s senior vice president of AI Risk and Advisory says they increasingly want to know whether a company keeps an inventory of its AI tools and use cases and who is responsible for each one. Some carriers use supplemental applications to collect this. The Lockton executive, who works in its Cyber & Technology business, says agent-to-agent interactions speed up transactions and reduce the human role in managing risk. Affirmative AI endorsements are growing in cyber insurance, though the executive says their value depends on the underlying policy wording, and standalone AI liability cover has not seen much uptake, partly because traditional technology E&O coverage is flexible. The executive argues that companies should assess, mitigate, and contractually transfer risk before turning to insurance.
News: States file 1,500 AI bills as insurers adapt
The brief: The National Conference of State Legislatures reports more than 1,500 AI-related bills introduced in 2026 so far, with 280 adopted, compared with more than 1,200 introduced and 227 enacted or adopted last year. Proposals requiring disclosure when people interact with AI nearly tripled this year, and deepfake legislation is the most common category of enacted AI laws. Wilson Elser’s privacy, security and AI practice leader compares the state-by-state patchwork to data breach and privacy laws. Resilience’s global director of insurance product for cyber says insurers are emphasizing risk selection and drafting AI policy language toward the EU AI Act’s standards. The article also notes the Justice Department’s AI Litigation Task Force, set up in January to challenge state AI laws it considers inconsistent with the national AI policy framework sought in a December 2025 executive order.
News: Panelists urge insurers to aim AI at better decisions
The brief: Panelists at the International Risk Management and Insurance Construction Risk Conference said builders and insurers adopting AI should prioritize better decisions and lower risk over efficiency. Zurich North America’s chief operating officer for global construction and surety said a one-point cut in loss ratio matters far more than a 10% cut in expenses, and that Zurich is restructuring its SharePoint files to improve the reliability of AI answers. Axa XL’s head of construction innovation and sustainability suggested AI can act as a fact-checker or a second pair of eyes in safety reviews so humans aren’t overloaded. The president of Marsh Risk U.S. and Canada said AI can help with project planning, progress monitoring, and early identification of challenges. The Zurich executive added that AI should empower workers rather than focus on efficiency alone.
