AI Liability Insurance Developments: A Shifting Scenery
In the rapidly evolving world of artificial intelligence (AI), businesses are increasingly relying on AI technologies for high-stakes, public-facing tasks such as customer response, recommendations, automated decision-making, and document drafting. However, as AI systems generate responses based on patterns learned from large data sets, outputs can vary significantly with similar inputs, leading to potential risks and liabilities.
These risks, including damage from hallucinations, model drift, mechanical failures, and other deviations from expected behaviour, have sparked a need for businesses to mitigate AI risks for their ongoing sustainability. One way this is being addressed is through the adoption of AI-specific insurance products.
Some insurance companies have started offering policies specifically designed to protect against AI-related risks, known as Affirmative AI-Specific Coverages. These policies aim to cover unique exposures that traditional insurance might not fully address. Additionally, existing lines of insurance, such as commercial general liability and directors and officers liability, can provide coverage for AI-related risks, albeit with the development of new policies to better address specific AI risks.
Insurers are also introducing exclusions to clarify what is not covered under traditional policies, known as AI Risk Exclusions. These exclusions help manage the unique risks associated with AI. Furthermore, some insurers are offering endorsements to traditional policies to include coverage for AI-related risks, known as Add-Ons for Traditional Policies. These endorsements can enhance existing policies to better protect businesses from AI-specific liabilities.
Cyber insurance may also cover AI liabilities related to data security, privacy, and network failures, including unauthorized access, data breaches, ransomware attacks, and legal or regulatory consequences. In the AI context, cyber insurance may apply when an AI system exposes personal or sensitive data due to a security flaw or unauthorized use.
However, it's important to note that most cyber policies do not contain express exclusions for AI-related incidents. For instance, Coalition, a cyber insurance company, expanded the definition of a security failure or data breach to include an "AI security event" and "expands the trigger for a funds transfer fraud (FTF) event to include fraudulent instruction transmitted through the use of deepfakes" or other AI technology.
As AI technology continues to evolve, the insurance industry is likely to continue adapting with new products and endorsements to meet emerging risks. For example, Armilla, a startup backed by Lloyd's of London, has developed a dedicated insurance product to cover financial losses tied to underperforming or malfunctioning AI models. AXA recently released an endorsement for its cyber policies that directly addresses risks associated with generative AI, stipulating coverage for a "machine learning wrongful act."
It's crucial for businesses to understand the potential risks associated with AI and the insurance products available to mitigate these risks. Directors and Officers (D&O) insurance provides coverage for claims made against a company's director, officer, employee, or the company itself based on mismanagement, breach of fiduciary duty, or violations of securities laws. As AI systems can be inconsistent, biased, or outright incorrect, leading to real legal, financial, and personal consequences, having the right insurance coverage can provide businesses with peace of mind as they navigate the complex world of AI.
Machine learning risks, such as model drift and data security issues, are being mitigated by businesses through the adoption of AI-specific insurance products like Affirmative AI-Specific Coverages and Add-Ons for Traditional Policies. Additionally, cyber insurance can provide protection against AI liabilities related to data privacy, network failures, and unauthorized access.
