AI Insurance: A Necessity Amidst Evolving Risks
As businesses increasingly integrate artificial intelligence (AI) into their operations, the question arises: what happens when AI goes wrong? The growing reliance on AI technology has led many companies, both large and small, to consider insurance as a means to manage the associated risks. However, acquiring coverage for these risks is not a straightforward process.
The Challenges of Obtaining AI Insurance
Many insurers exhibit hesitancy when it comes to covering AI risks, with some policies outright excluding AI coverage. This presents a dilemma for businesses, as operating without AI insurance could expose them to significant liabilities.
Compounding this issue is the rapidly changing and fragmented AI regulatory landscape. For example, California mandates reporting requirements for AI developers, while Colorado focuses on algorithmic discrimination. Tennessee regulates impersonation through voice and image, and various agencies, such as the FCC and FDA, have issued rules concerning AI applications. Additionally, Executive Order 14365 aims to streamline these state policies.
Recognizing the Risks of AI
As companies expand their AI applications, they become acutely aware of the risks involved. Major corporations are increasingly listing AI as a risk factor in their Form 10-Ks. While insurers acknowledge the risks presented by AI, they also recognize the opportunities it offers, as many are employing AI technologies themselves to enhance claims processing, underwriting, and fraud detection.
Litigation and Liability
The legal landscape surrounding AI is fraught with complexities. Cases such as Lokken v. UnitedHealth and Mobley v. Workday, Inc. highlight how AI can influence decision-making in healthcare and employment, respectively. Additionally, Raine v. OpenAI, Inc. raises concerns about product safety linked to chatbot outputs. The unpredictability of litigation outcomes, as seen in cases involving copyright issues, further complicates the insurability of AI technologies.
Factors Influencing Insurability
Insurability hinges on several principles: risks must be pure, quantifiable, fortuitous, and measurable. However, many AI applications do not readily fit these criteria. A prevalent issue is the “black box” phenomenon, where AI models produce outcomes in ways that are not comprehensible to humans. This poses challenges in high-stakes fields like healthcare, finance, and criminal justice.
Some insurers have opted for absolute exclusions regarding AI, introducing clauses that explicitly state that policies do not cover claims arising from AI-related technologies. Conversely, a market for tailored AI insurance has emerged since late 2018, with some insurers offering specific coverage to address unique AI risks.
Alternative Coverage Approaches
For those unable to secure tailored AI policies, “silent coverage” may be an option. This involves existing policies covering AI-related incidents without explicitly stating so. However, this approach carries risks, as it may leave certain incidents uncovered until claims arise. Another avenue is to acquire an algorithmic rider, which modifies existing policies to include specific AI coverage.
Navigating AI Insurability
To navigate the complex landscape of AI insurability, companies should focus on AI tools that align with accepted insurance frameworks. For instance, AI categorization and evaluation tools that meet established performance standards may be more readily insurable.
In contrast, generative AI poses unique challenges, as it creates outputs based on user prompts rather than analyzing existing data. For example, consider a hypothetical company that markets “Fido the Talking Dog,” where the generative AI used to create conversations could lead to unpredictable risks and liabilities.
Steps for Securing AI Coverage
Before implementing AI tools, companies should:
- Understand the scope of existing coverage.
- Consider adding an AI rider to current policies.
- Explore AI-specific insurance options.
- Incorporate insurance and regulatory requirements into compliance plans.
Conclusion: A Competitive Advantage
In an era where AI risks are prevalent, evaluating the need for AI insurance becomes crucial for businesses. Implementing robust compliance programs can provide a competitive edge, particularly as insurers navigate the complexities of AI coverage.
Key best practices include:
- Staying informed about applicable AI laws and regulations.
- Regularly evaluating insurance policies.
- Maintaining open communication with insurance providers.
- Conducting audits of AI tools before integration.
- Adding indemnification provisions to agreements with AI providers.
- Developing comprehensive AI compliance plans.
As the demand for AI insurance grows, businesses must adopt proactive strategies to mitigate risks and secure appropriate coverage in this evolving landscape.
