Insurers Urged to Keep Human Oversight Central as Agentic AI Adoption Grows

Insurers are beginning to deploy agentic artificial intelligence to automate underwriting, document verification, and customer servicing, but human oversight must remain central to decisions that carry financial and regulatory risk, according to German IT consulting firm valantic. Speaking at the Asian Banking & Finance and Insurance Asia Summit in Singapore, valantic APAC Director Goncalo Carvalho demonstrated a life insurance workflow where AI agents independently verified customer identification, proof of address, and medical history before routing applications to human underwriters.

Carvalho noted that enterprises are expected to increase AI projects by 2,500% by 2028, yet 67% of IT leaders remain uncomfortable with the technology. His demonstration showed how agentic AI can generate risk assessments and allow underwriters to query the reasoning behind recommendations through a chatbot interface — but the final decision stayed with a human. Acting as the underwriter, Carvalho overrode the AI’s recommendation and approved an application manually, illustrating the principle that accountability cannot be delegated.

“With AI, development is becoming much easier. But the important question is: are we building solutions that are safe, secure, scalable, and governed properly?” Carvalho said. The workflow automatically generated audit logs recording whether each action was taken by an AI agent or a human user — a critical feature for regulators as Singapore’s financial sector grapples with how to supervise AI-driven decision-making.

The presentation comes as Singapore’s insurance sector faces growing pressure to adopt AI while navigating a tightening regulatory environment. The Monetary Authority of Singapore recently proposed the SAFR framework to govern AI agents in finance, signalling that agentic AI — which can execute multi-step processes autonomously — will face increased scrutiny. Singapore-based insurers including Etiqa have already deployed AI chatbots for policy inquiries, and the shift toward autonomous AI agents represents the next frontier.

Why it matters for Singapore: As a regional insurance hub, Singapore’s approach to agentic AI governance will set precedents for the broader APAC market. The balance Carvalho described — between AI-powered efficiency and human accountability — maps directly onto MAS’s emerging regulatory philosophy of enabling innovation while mandating guardrails. Insurers operating out of Singapore would do well to implement audit trails and override mechanisms now, before compliance becomes a bottleneck.