Gallagher’s third annual Artificial Intelligence Adoption and Risk Survey shows that businesses are moving Artificial Intelligence from pilot projects into daily operations faster than they are building governance around it. Based on responses from more than 1,200 global businesses, the survey found that nearly two-thirds (62%) have delivered Artificial Intelligence training to employees in the last year, and more than half (55%) have hired for Artificial Intelligence-focused roles. Governance is becoming more visible internally, with 56% of organisations already communicating an Artificial Intelligence strategy to employees. However, 43% are yet to introduce a formal Artificial Intelligence risk management framework, and only 44% have conducted Artificial Intelligence impact assessments.
The findings point to a shift from experimentation to workforce enablement. Almost half of businesses (47%) now offer training to help employees use Artificial Intelligence tools, up seven percentage points from 2024. Four in 10 (40%) have created new roles where Artificial Intelligence is a core part of the remit, reflecting demand for data, automation and Artificial Intelligence governance skills. The vast majority of respondents (86%) said Artificial Intelligence has improved employee productivity, with use focused on repetitive tasks such as document handling, triage, underwriting workbenches and customer service rather than full automation of complex underwriting or claims decisions. At the same time, businesses continue to see human capability as central, with 34% citing the need to preserve a human touch in client interactions and 31% saying people are still needed to solve complex problems technology cannot yet address independently.
For insurers, the governance gap is feeding into more complex technology E&O, cyber, D&O and employment practices exposures as Artificial Intelligence tools are embedded in underwriting, pricing, claims, human resources and customer interactions. Regulators and courts are also moving toward more concrete expectations. In the EU, the Artificial Intelligence Act is being phased in from 2024 to 2026, while in North America, supervisors including the NAIC in the US and OSFI in Canada have issued guidance on model risk management, data governance and the use of Artificial Intelligence and big data in underwriting and claims. Gallagher said disputes are already emerging over algorithmic failures, data leakage from training sets and Artificial Intelligence-assisted social engineering, while directors and officers underwriters are beginning to ask how boards oversee Artificial Intelligence strategy and risk.
The research also points to a broader advisory opportunity for brokers and risk consultants. As clients expand Artificial Intelligence training and hiring without matching investment in oversight, intermediaries are positioned to help build risk frameworks, map controls to new use cases and prepare for disclosure expectations. Over time, that is likely to lead to more detailed Artificial Intelligence-specific questions at placement and renewal, including which processes are algorithm-driven, how models are validated, how vendors are managed and how complaints or errors are handled. Gallagher’s view is that long-term value will depend on combining technological efficiency with human creativity, judgment and trust.
