Gallagher Re issued a report warning that current AI model evaluation methods are inadequate for underwriting AI risk, emphasizing the need for improved assessment focusing on model failures and correlated risks. The report critiques reliance on benchmark performance tests and highlights challenges from restricted-distribution models like Anthropic’s Mythos, which limit independent evaluation. This affects insurers’ ability to price AI risk accurately, potentially increasing uncertainty loadings, underwriting costs, and concentration risk in the commercial insurance market.
Read Source arrow_forward
QBE reported that 50% of Australian businesses experienced a cyber incident in the past year, with 26% of these involving artificial intelligence. The survey identified AI-driven attacks such as vulnerability identification, malware, phishing, deepfakes, and business email compromise, and showed that 66% of incidents were linked to third-party suppliers, highlighting interconnected risk exposure. This development increases cyber risk complexity, leading to greater demand for cyber insurance and driving higher IT security spending among commercial brokers and carriers to manage operational and supply chain vulnerabilities.
Read Source arrow_forward
ePayPolicy reported that over 6,000 insurance industry leaders convened at Insurtech Insights USA to discuss how legacy systems and AI-driven modernization are shaping the future of insurance. The event highlighted that insurers struggle with outdated infrastructure that limits integration and agility, while AI and connected APIs offer tools to automate workflows, improve underwriting precision, and enhance customer experience. These shifts affect commercial broker margins and carrier efficiency by emphasizing incremental modernization over large-scale projects and aligning technology upgrades with customer and operational needs.
Read Source arrow_forward
Guillaume Bonnissent, CEO of Quotech, discussed the challenges of securing confidential data when using large language model (LLM) chatbots, including differences between so-called "enterprise" AI accounts and cloud platforms like AWS Bedrock. The article explains that Bedrock provides data processing within a customer’s cloud tenancy to prevent data from being used in AI training, while typical enterprise AI accounts may not guarantee data protection without reviewing contracts such as DPAs or BAAs. This distinction matters because secure data handling and jurisdictional control impact insurer and MGA compliance, operational risk, and potential data leakage that could affect underwriting efficiency and broker risk management.
Read Source arrow_forward
The UK government's online safety crackdown is shifting liability exposure from social media content to product design risks, affecting policies traditionally focused on media, privacy, and cyber coverage. The change involves legal claims centered on defective design, failure to warn, and regulatory compliance tied to platform and device architecture rather than user-generated content. This shift increases insurer exposure to product liability, recall risk, and regulatory fines, complicating underwriting and potentially reducing commercial broker margins on technology-related insurance products.
Read Source arrow_forward
Sixfold announced the launch of its AI Underwriter, an AI agent designed to provide straight-through quote and bind capabilities for property and casualty insurers, supported by $30 million in Series B funding completed in January 2026. The AI Underwriter integrates decisions from underwriters into a proprietary institutional memory, maintaining strict data separation between carriers and allowing configurable autonomous processing up to quote-ready and bind-ready outputs. This technology decreases underwriting processing times by 50% to 97%, increases hit ratios by 15% or more, and raises gross written premium per underwriter by up to 30%, improving carrier underwriting efficiency and capacity.
Read Source arrow_forward
The Insurance Information Institute (Triple-I) and Munich Re released the RiskScan 2026 survey based on responses from over 1,700 participants, identifying cyber incidents, economic pressures, and AI as top insurance risks. The survey used online data collection by RTi Research to assess risk exposures in the U.S. and U.K., analyzing factors like inflation, geopolitical uncertainty, and supply chain issues as risk multipliers. These findings highlight increased challenges to insurer capital allocation, underwriting accuracy, and market stability, influencing commercial broker margins and carrier risk management strategies.
Read Source arrow_forward