Artificial intelligence is transforming industries at an unprecedented pace, and the insurance industry is trying to keep up.

As AI integrates into business operations, from customer service to facilities management to supply and distribution to internal processes and business management, the risks associated with AI increase dramatically. And as risk increases, so too does insurer scrutiny. Such is becoming the case with AI.

This article explores the unique challenges in assessing AI risk particularly when it comes to answering questions about AI on insurance applications, and potential best practices for mitigating associated risks.

AI Questions on Insurance Applications

As businesses integrate AI into their operations, insurers are starting to develop targeted questions to assess the associated risks. These inquiries often focus on how a company may be using AI and its impact on business functions and revenue.

As one example, an application from Philadelphia Indemnity Insurance Co. requires applicants to: (1) estimate the percentage of revenue derived from "artificial intelligence software/services," and (2) disclose whether their business uses "Generative AI in producing original content," including for advertising or branding.

While these questions may seem straightforward, answering them accurately can be difficult for businesses given the complexities of AI. For example, the term "artificial intelligence" potentially encompasses various technologies, including machine learning, natural language processing and generative AI.

The Philadelphia Indemnity application, however, does not supply a meaningful definition of "generative AI," or even "AI." Without a clear definition in the application itself, businesses may struggle to determine which operations even qualify as AI for purposes of responding to the insurer's questions.

Insurers may also fail to specify critical parameters such as time frames for reporting AI usage or whether they seek disclosures about third-party AI systems, services or practices. These ambiguities, combined with the rapid evolution of AI technologies, which may cause even the most accurate responses to quickly become stale, can create challenges for businesses applying for insurance.

Importance of Accurate Disclosures

In this dynamic environment, the stakes for accuracy in disclosures is high. Depending on how disclosed information bears on a future loss, an incomplete or inaccurate response about the prospective insured's use of AI could result in claim denial or, worse, policy rescission. Given these risks, businesses should adhere closely to best practices when facing renewal or application questions concerning AI.

Define AI in a business-specific context.

Given the breadth of the term "artificial intelligence" and the lack of any universally accepted definition, businesses should work with their insurers and brokers to reach a meaningful and functional definition of AI. The agreed definition can be supplied by addendum to the application and can specify the types of technologies and processes that will be considered AI for purposes of that particular business.

This proactive clarification can help align the business's understanding with the insurer's understanding, potentially minimizing coverage disputes.

Engage all stakeholders.

Unlike commonplace risks like fire and flood, and liabilities like environmental, products, employment and management, understanding and quantifying the risk that AI poses to a particular business is likely beyond the reach of dedicated risk management personnel.

To fully appreciate the risk posed by AI requires input from all aspects of a business. For instance, human resources may use AI to screen employees or monitor behaviors, supply and distribution may use AI to assist with product and product routing, production may use AI across its automated facilities and production lines, and management may use AI to provide strategic advice.

In fact, earlier this year it was announced that a company in the United Arab Emirates, International Holding Co., appointed an AI-powered observer to its board of directors. This follows news that a Finnish company, Tietoevry, named an AI entity to its leadership team. A Hong Kong-based company, Deep Knowledge Ventures, appointed a computer algorithm to its board of directors in 2014.

Given these broad and varied uses of AI, it is important for businesses to engage all units and stakeholders when it comes to assessing the company's use of AI and, thus, the risk posed by AI.

Clarify third-party AI usage limitations.

Many businesses rely on third-party vendors or partners that incorporate AI into their services. Since businesses often lack full visibility into these third parties' internal operations, businesses may wish to clarify this limitation when answering questions about AI on insurance applications. That is, prospective insureds should make reasonable inquiry to third-party vendors and partners about their use of AI and include such information on their insurance application.

Where a company is unable to obtain the information after reasonable inquiry, it should so state. This could protect the business from potential coverage disputes should undisclosed third-party AI cause or contribute to a loss.

Provide date-specific information. 

Given the rapid pace of AI innovation, businesses should cabin AI-related disclosures to a specific time frame. In doing so, businesses can avoid accusations that later changes to AI systems or processes render their prior disclosures inaccurate or misleading. In addition, businesses should be explicit about any duties to periodically update the insurer regarding the use of AI during the policy period.

Conclusion

As businesses become more dependent upon the use of AI, insurers are certain to increase their inquiries about it when it comes to the procurement of insurance. Businesses should be on the lookout for AI policy endorsements, provisions and, importantly, questions about AI in insurance applications.

As with all insurance application information, accuracy about AI is critical and businesses, therefore, must ensure that their insurance procurement team is fully informed about the business's use of AI and adhere closely to industry best practices when making any AI disclosures.

---

The opinions expressed are those of the author(s) and do not necessarily reflect the views of the firm, its clients, or Portfolio Media Inc., or any of its or their respective affiliates. This article is for general information purposes and is not intended to be and should not be taken as legal advice.