Artificial intelligence (AI) continues to reshape the way businesses operate, from human resources and operational efficiency to cybersecurity and financial reporting. It should come as no surprise, therefore, that companies are calling on AI to facilitate and enhance corporate filings and shareholder communications. Management Discussion and Analysis (MD&A) submissions in publicly-traded companies’ Securities and Exchange Commission filings are no exception. These submissions have been viewed by securities analysts as a sort of corporate DNA which, if read properly, could reveal telling traits and warning signs about future corporate performance. While it might be common knowledge that analysts are using AI to analyze MD&A submissions, less clear is whether (and which) companies are using AI to generate their MD&As.

One of the greatest and most publicized “flaws” with AI is the technology’s tendency to hallucinate, by generating seemingly plausible but ultimately false or misleading outputs. And while some hallucinations may be readily obvious, others, particularly those based on trends, where the output is predicting a result based on certain known or assumed criteria, may be more difficult to recognize with the naked eye. Therefore, without rigorous human oversight, it is easy to see how a hallucination might end up in a filed corporate disclosure and thereby run afoul of SEC rules prohibiting false or misleading statements about material facts.

Inaccuracies or misstatements in securities filings are frequently the root of securities litigation. An increase in AI-generated MD&A outputs can lead to increased securities claims if companies do not adequately prepare to identify and mitigate the risk of inaccurate disclosures.

The costs associated with securities litigations are typically covered by directors and officers (D&O) liability insurance, which offers coverage for lawsuits arising out of alleged securities violations, including shareholder derivative lawsuits related to misleading disclosures. But not all policies are created equal, and the availability and scope of coverage for AI-related exposures, including in securities claims, can vary widely. Companies using AI to generate the contents of their MD&A sections would be well served to ensure their current liability insurance program provides adequate coverage in the event AI-generated content in an MD&A results in a claim.

In a recent recent post, we discussed the growing number of AI exclusions in insurance policies, and specifically in D&O and E&O policies. Depending on the breadth of the exclusion, even a seemingly narrow AI exclusion could, in reality, bar coverage for a routinely covered securities claim or suit if the suit is predicated on information supplied or refined with AI. Similarly, broader exclusions might impact claims where even the data considered in the MD&A is the product of AI use in corporate operations.

Companies should assess their D&O programs to minimize or eliminate unintended coverage gaps for AI risks. Key considerations include:

1. Policy Review: Ensuring that AI-related losses are covered and not subject to exclusions, including for cybersecurity, technology, or AI risks. Analyzing insurance programs in their entirety and paying close attention to mid-year policy changes or endorsements can help policyholders identify potential oversights in their current insurance program and bolster complementary coverages for AI risks.

2. Regulatory Coverage: Verifying that policies provide coverage not only for shareholder claims but also for regulator claims and government investigations that may arise from misstatements, errors, or omissions in securities filings or disclosures. Securing investigation coverage (if available), enhancing its scope, and addressing problematic sublimits or limitations can help guard against regulatory risks.

3. Consider Purchasing AI-Specific Coverage: Affirmative AI insurance coverage continues to proliferate in the form of standalone AI insurance policies and endorsements. Such coverage can offer additional protection for companies who utilize AI in their operations.

4. Consider a Chief AI Officer, and Ensure They Are Covered: Just as the widespread use of technology led to the CISO, companies are now considering similar positions for AI. AI is quickly becoming ubiquitous in technology platforms, manufacturing systems, building automation, communication, logistics and supply chain, along with virtually every other aspect of life. Understanding how a company is using AI, where it is using it, and what potential risks it brings, requires a dedicated person or team. This is especially the case when it comes to corporate statements and disclosures. In fact, conversations with insurance underwriters suggest that we may soon see requirements that a C-suite level officer “sign-off” on corporate statements as a condition to coverage for AI risk.

As we’ve discussed previously, adding new members to the C-suite can also raise D&O coverage issues, including whether those individuals are covered “insureds” in the first place. With personal liability risks for directors and officers continuing to evolve, companies must also adapt their D&O programs to recruit, retain, and protect top talent.

5. Consider AI Training and Educational Initiatives: Ensuring that key stakeholders such as directors, officers, and other key stakeholders are familiar with the AI technologies that a company uses can help companies mitigate AI risks. Familiarity with AI technologies can also ensure that AI-created mistakes in documents such as a company’s MD&A do not go undetected and can be rectified long before filing.

As businesses increasingly use AI to generate and analyze content for regulatory disclosures like MD&A sections, they face heightened risks of unintended errors that can lead to securities litigation. While (D&O) liability insurance typically covers such claims, the growing prevalence of AI exclusions in D&O and other coverage lines means policyholders must carefully and regularly assess their insurance programs to avoid accidental pitfalls in their coverage. Experienced coverage counsel, insurance brokers, and other risk professionals can assist policyholders in navigating this ever-changing risk landscape.