Free Preview

This is a preview of some of our exclusive, member only content. If you enjoy this article, please consider becoming a member.

Financial services technologies come and go. Useful shiny objects become standard operating procedure. How quickly and effectively an insurer makes this transition varies. Faster and better yields more business value.

Executive Summary

Banks learned the hard way—with online banking—that setting up separate, standalone organizational structures is an inefficient and costly way of propagating new technologies. SAS Institute’s Insurance Industry Advisor Mike Fitzgerald advises insurers not to make the same mistake with AI, urging them to include AI considerations in existing processes.

Here, he offers a logical place to start—the board of directors—outlining some of the changes AI-powered insurers will make to standing committees of their boards to address the opportunities and risks associated with AI.

Online banking is a classic case. Many banks initially set up their online banking as a separate, standalone unit. This was costly as it required parallel management structures, duplicated existing business processes and localized knowledge about digital banking in a specific group of employees.

With time, online became part of the core banking business and just another “way we do business around here.” Costs were reduced and processes streamlined.

Insurers that use existing organization structures to propagate new technologies are examples of faster and better. AI is no exception. Insurers have an opportunity to speed and improve AI implementations by including AI considerations into existing processes.

The place to start is the standing committees of the boards of directors. This article outlines the adjustments that AI-powered insurers are making to their board committees in order to address both the opportunities and risks of AI.

Board Committees

A board of directors sets an insurer’s strategy and policies, hires and oversees the performance of the company’s executive officers, and ensures that the company complies with legal and regulatory requirements.

Importantly, the board is the last line of defense for the brand of the insurer. It ensures that its strategy and values further the agreed purpose of the organization.

“AI has the potential to bolster or destroy a brand.”

AI has the potential to bolster or destroy a brand. Boards are working to protect their firms while harvesting full value from the tools. To accomplish both goals, they will adjust their existing committee structure.

Insurance board committees vary but usually include groups that deal with nominating and governance, compliance, audit, risk, investment, underwriting, investment, and compensation. The key committees involved in AI implementation will make the following adjustments:

  • Nominating and Governance

This committee recruits board members. It identifies individuals qualified to become directors and recommends nominees for the next annual meeting of shareholders. These nominees are traditionally recognized experts in insurance (underwriting and claims), finance and legal. In the past, skillsets around technology were not a focus of recruitment.

Going forward, recruiting board members that understand AI to a level that allows them to make sound strategic decisions will be necessary. Nominating and governance committees in AI-powered insurers will add recruitment of individuals with these skills as a priority.

  • Compliance

Compliance committees monitor the regulatory environment, review compliance reports, and oversee the company’s ethics and compliance programs. Traditionally, this has been around addressing the aggregated results of market conduct studies, responding to existing and pending insurance legislation, and funding operational compliance areas.

Recently, insurance regulators in multiple jurisdictions have been taking action on AI. In the U.S., the NAIC established working groups on big data and artificial intelligence; in Canada the Office of the Superintendent of Financial Institutions (OSFI) published its expectations around model risk management; in Europe, the EU Parliament is in final discussions on its Artificial Intelligence Act.

(Editor’s Note: The NAIC adopted a model bulletin on AI in December, after this article was first published)

Board compliance committees will add AI regulation to the list of laws they monitor and evaluate. Their work will inform other committees what needs to be done and also what may be coming in the future.

  • Audit

This committee works with internal and external auditors to ensure standards are met across the organization. They establish and update their expectations for executive management to ensure that the risks of the organization are identified, eliminated and mitigated. When effective, their reach extends to every critical part of the organization without being over burdensome or bureaucratic.

Existing standards related to the integrity of the firm and compliance with laws and regulations will be updated to include AI. For example, the audit committee will publish expectations concerning the frequency and depth of AI use audits, data security and privacy reviews, and legal agreements with AI providers.

  • Risk

Artificial intelligence democratizes data access, allowing unprecedented access to data that was previously guarded by programmers, data scientists and other technicians. This exposes an insurer to reputational, liability, bias/discrimination, ethical, intellectual property and trust risks.

The risk committee has existing processes to measure and monitor corporate risks. For AI, the committee will use these same processes to ask management to report on the type and extent of AI use, measure the risks involved with this use, and provide updates to them on risk management activities related to AI exposures.

  • Underwriting

Underwriting committees serve as a critical oversight and governance function, ensuring that the company’s underwriting activities are conducted in a prudent and responsible manner and that underwriting risks are managed effectively. Approving and monitoring underwriting guidelines, tracking underwriting risk exposures, and approval of large or complex risks and reinsurance treaties are common tasks.

“Boards are working to protect their firms while harvesting full value from AI tools. To accomplish both goals, they will adjust their existing committee structure.”

Going forward, this committee must ensure the intentional use of AI in underwriting. Does the firm want to use AI only to recommend actions? Where and when should this be extended to underwriting decisioning? Are there lines of business for which AI will not be used, or lines where AI application should be aggressively pursued? The committee will review and approve AI underwriting guidelines, evaluate AI underwriting performance and ensure that AI underwriting activities are aligned with the company’s risk appetite.

This is a great deal of change for the board. Every insurer has its own experience to draw on when it comes to successfully making such broad change happen. One approach involves the appointment of a leader by the chairman and CEO who would work with the committees and adjust their operating models to incorporate AI. As these committees then begin to ask carrier operating units to provide input into their deliberations, this same leader would assist the operating units to satisfy the board’s request. For example, the underwriting committee will ask the chief underwriter officer to provide them with the AI underwriting guidelines. Through this process, AI is incorporated into the business process through the insurer’s existing governance and control procedures.

Including AI in existing board committee processes will not eradicate the AI learning curve. But it should flatten it a bit and ensure that effective results are achieved faster and at lower cost.