Goldman: 60% AI Economics by 2030

The slow scaling of AI in finance is not for a lack of potential return on investment; commercial property and casualty insurers using agentic AI have seen loss ratio improvements of 3-5 percentage points and a staggering 60-99% reduction in the time it takes to get a quote. Instead, the sluggishness is attributed to a combination of factors including a shortage of specialized AI talent, the complexity of integrating with legacy systems, and concerns around data quality and governance. While a 2024 survey showed the insurance industry is outpacing most other sectors in the early adoption and testing of AI, the reality is that the majority of these initiatives are still in the experimental phase. Approximately two-thirds of insurers remain in the piloting stage, with only about 7% having successfully implemented AI systems at a large scale across their organizations. Similarly, while 76% of U.S. insurers have implemented generative AI in at least one business function, a mere 10% have achieved widespread deployment. Agentic AI, which can act autonomously to achieve goals, is a key focus within the industry. In underwriting, these AI agents can independently handle the entire workflow, from ingesting and structuring data from broker submissions to generating quotes and supporting negotiations. This can lead to a 30-60% reduction in cycle time and allows underwriters to focus on more complex risks. In claims processing, agentic AI is being used to automate intake, validate policies, triage the severity of the claim, and even initiate the first steps of the settlement process. For instance, Allianz has launched "Project Nemo," an agentic AI solution that automates food spoilage claims, cutting processing times from days down to just hours. This system utilizes seven specialized AI agents to manage tasks from coverage verification to fraud detection, with a human always making the final decision on payment. The primary barriers to broader AI adoption in the insurance sector include the high costs of implementation, a lack of technical expertise, and navigating regulatory and compliance issues. Many employees in the insurance sector lack the necessary skills to effectively label and interpret the data generated by AI systems. Furthermore, there are ethical considerations, as AI systems trained on historical data could perpetuate existing biases. Looking ahead, the successful implementation of AI will require significant investment in training programs to upskill the current workforce. For those companies that successfully scale their AI initiatives, the competitive advantages are substantial. Leaders in AI adoption within the insurance industry have already demonstrated a total shareholder return that is 6.1 times higher than that of their slower-moving competitors over a five-year period. The move towards more sophisticated AI is also evident in the fintech space, with major payment processors like Visa and Mastercard developing agentic AI payment solutions. These systems aim to embed intelligent, autonomous payment capabilities directly into commerce platforms. In the broader financial services sector, companies like JPMorgan Chase are exploring the use of AI agents for fraud detection, personalized financial advice, and the automation of loan approvals.