New study: Ethical and regulatory concerns remain high, despite 9 in 10 insurers planning to invest in GenAI in the next year 

Nearly 250 insurance decision makers share their outlook on GenAI strategy, sentiment and innovation

A new study exploring the use of generative AI (GenAI) in insurance suggests that 9 in 10 insurers plan to invest in GenAI in the next year, but the ethical and regulatory implications of innovation continue to frustrate many.

The study, Your journey to a GenAI future: An insurer’s strategic path to success, comes from a global multi-industry survey by SAS, a data and AI company serving 47 of the world’s 50 largest global insurers, and Coleman Parks Research Ltd. Its findings offer an insider’s view on how insurance companies worldwide are implementing, budgeting for and developing strategies around GenAI, based on survey insights from 236 industry decision makers.

With 89% of insurance sector respondents planning to invest in GenAI in 2025, the research found that 92% of that number have a dedicated GenAI budget in the works.

Among the industry’s goals for investing in GenAI, the top three emerged as:

1. Improvement in customer satisfaction and retention (81%, the highest of any industry segment).
2. Reduction in operational costs and time savings (76%).
3. Enhanced risk management and compliance measures (72%).
    

Already, two-thirds (68%) of insurance professionals surveyed reported using some form of GenAI in their professional roles at least once a week. About 1 in 5 (22%) professed they use the technology daily. While only 11% of respondents said their organisation had fully implemented GenAI, another 49% indicated they were already in the process of implementing it.

However, insurance decision makers showed themselves modestly more anxious about GenAI ethics than their counterparts in other industries. Among insurance respondents, 59% indicated concern about the ethical implications of their organisation’s GenAI, in contrast to a cross-industry average of 52%.

Despite insurers’ deeper ethics worries, their plans for governance and monitoring – efforts that would include the creation, implementation and maintenance of ethical frameworks – remain a work in progress:

• Only 5% of insurance respondents described their organisation’s GenAI governance framework as “well-established and comprehensive”.
• 57% reported that their organisation’s frameworks were “in development”.
• 27% called their organisation’s frameworks “ad hoc or informal”.
• 11% said their ethical frameworks were “non-existent”.

Complementing concerns about AI ethics are regulatory compliance worries., with only 1 in 10 (11%) of insurance respondents reporting their organisation is fully prepared to comply with current and upcoming GenAI regulations.

Commenting on the research, Joe Rowe, Data and AI Insurance Lead for UK, Ireland and Africa at Accenture, said:

“GenAI is not a silver bullet, but insurers are finding it can provide many more pieces of the jigsaw puzzle, including in areas that have previously proven quite difficult, like the ingestion of unstructured data. Claims and underwriting are prime examples where GenAI is helping the human in the loop extract insights and make better decisions.

“The use of GenAI is progressing quite rapidly, but to develop it responsibly, insurers must have an alignment of people, processes and technology, all working together to drive use cases from experimentation into operations and production. Proper governance requires focus and investment.”

Andrew Pollard, insurance specialist at SAS UK and Ireland, added:

"While insurance is often seen as a slow-moving industry, insurers are clearly emerging as frontrunners in the GenAI space, shown by significant investment and enthusiasm. To unlock its potential, however, responsible innovation must be embraced and strong policies and processes put in place to protect customers and ensure the integrity of data used in models.” 

Large language models (LLMs), for example, require huge amounts of data to properly treat edge cases, yet the insurance industry faces a "data drought" of large datasets, combed for bias and checked for data quality. This is crucial because the quality and quantity of data used to train GenAI and other AI models can make or break the accuracy, fairness and equity of the model’s results in claims and policy decisions.

Furthermore, insurers, as fiduciaries, safeguard significant volumes of sensitive personal identifiable information. With data privacy anxiety growing, synthetic data - artificial data manufactured to realistically mimic real-world data, used to enrich existing datasets without compromising customer privacy – could provide an answer.

More than a quarter (27%) of insurance industry survey respondents reported using synthetic data; nearly a third (30%) said they were actively considering it, and 22% reported they might consider it.

“The next step is clear”, added Pollard. “Insurers must embrace ethical frameworks and data rigour as their guiding principle  to realise the full value of GenAI technology.”

To find out more about this exclusive research from Coleman Parks and SAS,  watch a new webinar, Unleash the Power of Generative AI in Insurance, available now on demand, featuring experts from Microsoft, Accenture and FRG.

Additional insurance and cross-sector results from industries like banking, healthcare, life sciences and government can be compared and contrasted via SAS’ interactive GenAI survey data dashboard.

Visit SAS' GenAI data dashboard to explore how the insurance sector compares to other industries.

Study findings suggest that GenAI could prove to be a competitive differentiator for early adopters.