Skip to main content
Choose your location
Select the location and the language that you prefer
main content, press tab to continue
Article

A perfect fit: Generative artificial intelligence & corporate insurance

By Anthony Rapa | July 3, 2024

Knowing where to look for coverage is key to solving the AI puzzle.

Generative Artificial Intelligence (Gen AI) is gaining traction with financial institutions, transitioning from an emerging technology to a strategic tool. Most firms are now in the early stages of adoption, focusing on governance frameworks, evaluating risks, and understanding how Gen AI fits into their plans. Many initial Gen AI uses are taking place in the back office, where the technology helps drives efficiency through increased productivity and automation of repeatable tasks. In the mid to long run, expect Gen AI to play an increasingly important role in the front office, helping firms better engage with customers and to create new and innovative advisory and financial products.

Managing the risks associated with Gen AI is like solving a complex puzzle—no single insurance policy covers all potential exposures. Instead, a comprehensive risk management strategy must be pieced together using various policies from your insurance portfolio.

Unpacking the pieces

As a general-purpose technology, Gen AI can be integrated into a wide range of applications. Accordingly, Gen AI creates risks that can permeate potentially every facet of a financial institution’s operations, creating significant disruptions to existing risk profiles while creating new and potentially uncovered risks. In no particular order, those risks might include:

  1. 01

    Bias & Explainability

    Complex AI models makes it challenging to explain their decisions. AI systems may inadvertently reflect and perpetuate biases present in training data.


  2. 02

    Quality Concerns & Hallucination

    Produces plagiarized content due to limited training data or a lack of understanding of originality.


  3. 03

    Overreliance

    The growing capabilities of AI raise the concern of excessive reliance without sufficient critical evaluation.


  4. 04

    Infrastructure & Third Parties

    Companies relying on external parties may face risks associated with the reliability, security, and continuity of those services. Internal resources require investments in talent and data storage.


  5. 05

    Regulatory

    Rapid advancements in AI may outpace the development of appropriate regulations, leaving companies navigating uncertain legal landscapes.


  6. 06

    Data Privacy & IP

    AI's complex algorithms and data usage fuel IP and data privacy risks. Mass data usage challenges the definition and protection of rights, ownership, and liability.


  7. 07

    Crime

    Gen AI tools can be misused for social engineering attacks, enabling financial scams and the creation of sophisticated fake identities & deep fakes.


  8. 08

    Liability

    How to determine which party is at fault when Gen AI makes a decision?


  9. 09

    Content moderation

    The diversity and unpredictability of generated outputs makes it difficult to establish clear content moderation guidelines.


  10. 10

    Data accuracy

    Biased, incomplete, or erroneous data leading to inaccurate outcomes, amplifying misjudgments and perpetuating disparities in decision-making processes. Datapoisoning.


  11. 11

    Ethics

    AI applications may raise ethical dilemmas, such as the use of facial recognition, surveillance, or autonomous decision-making in sensitive areas like healthcare and criminal justice.


  12. 12

    Workforce Disruption

    As more companies adopt AI applications to automate their process, there is an increased risk of job displacements and unemployment.


Taking stock of the puzzle pieces

It's important to note that coverage is, as always, highly fact-dependent, with the likelihood of finding coverage varying based on the specific circumstances of each claim or scenario.

To manage the risks posed by Gen AI effectively, financial institutions should adopt a multifaceted approach to insurance. This involves understanding how different risks align with various insurance policies. Our heat map illustrates the correlation between some of the top Gen AI risks and many of the insurance policies carried by firms today. Boxes codded “green” represent a high likelihood of finding coverage; “red”, a low likelihood; and “yellow” where coverage will depend heavily on the particulars of the claim and policy language at issue.

Putting the pieces together

EPL insurance will play a crucial role in covering risks arising from these changes. As institutions restructure their workforce and implement new talent strategies, they may face allegations of wrongful termination, discrimination, or other employment-related issues. EPL insurance provides protection against such claims, ensuring that institutions can navigate these transitions with reduced financial risk.

Another significant consideration under EPL is third-party discrimination or harassment. Many overlook that EPL insurance covers claims of discrimination or harassment made by non-employees, such as customers or job applicants, even when these allegations are not directly related to employment within the firm. This coverage is vital in scenarios where Gen AI applications may inadvertently cause harm.

Example: A bank uses a Gen AI-powered chatbot to assist with customer service. If the chatbot displays bias against a protected class, violating anti-discrimination laws, the bank could face allegations of third-party discrimination. EPL insurance would help cover the legal costs and potential settlements associated with such claims.

Putting the pieces together

Managing the risks associated with Gen AI requires a comprehensive and holistic approach. Financial institutions must consider how each piece of their insurance coverage fits together to form a complete risk management strategy. Addressing the unique challenges posed by Gen AI across various policies helps protect against potential exposures.

Financial institutions may want to consider the following as part of that strategy:

Scenario analysis & crisis response planning:

  • Like a cyber tabletop, conduct scenario analysis and crisis response planning to reveal coverage gaps, necessary improvements, and integration of insurance into AI governance frameworks.

Cross-policy coordination:

  • Just because a risk is excluded under one policy doesn’t mean it’s automatically covered elsewhere. Ensure cross-policy coordination by scrutinizing coverage provisions, substantive wording, and other clauses, while considering different deductibles/retentions, limits, and notice provisions.

Question benchmarking:

  • As Gen AI use increases and the technology improves, mirroring your neighbor’s approach to insurance might not be all that useful. Avoid relying solely on benchmarking for limit adequacy and retention levels. Use advanced analytics to understand your firm’s unique risk profile and make better, bespoke risk management decisions.

Proactive underwriter engagement:

  • In case there was any doubt, financial institutions are now in 2024 being asked by underwriters about Gen AI. Consider getting ahead of these questions through proactive engagement, demonstrating your firm’s superior governance framework and understanding of your risks. While these factors might not drive outsized renewal results today, the market and the risks will continue to evolve. It’s not hard to imagine a world, not unlike the current cyber insurance environment, where strong Gen AI governance makes the difference between renewal success and heartache.

Scrutinize policy wording & proposed changes:

  • Although Gen AI isn’t currently excluded under most policies, that doesn’t necessarily mean that it’s covered, either. Gen AI presents several nuances that might frustrate existing policy wordings, potentially leading to claim complications or even outright denials. Work with a broker who understands your policies at a technical level to scrutinize possible Gen AI claim scenarios for your firm against existing policy wordings. Then, at each renewal, consider how proposed changes might impact your coverage for Gen AI and associated risks.

Conclusion

By taking a proactive approach, financial institutions can effectively piece together a robust insurance strategy that addresses the multifaceted risks of Gen AI. Just like solving a complex puzzle, the key to success lies in understanding how each piece fits into the bigger picture, ensuring comprehensive and cohesive coverage.

Author


FinTech Subvertical Leader, Financial Institutions & Professional Services – North America

Contacts


Medina El-Farra
Team Leader – FINEX Financial Institutions, Canada

Contact us