Insurers Are Now Covering AI Malfunction Losses

Uploaded on 2025-05-20 in TECHNOLOGY-Key Areas-Artificial Intelligence, NEWS-Cybersecurity News, FREE TO VIEW, BUSINESS-Services-Financial

Lloyd’s of London has recently launched an innovative insurance service designed to protect organisations from losses related to AI malfunctions. This initiative aims to capitalise on industry concerns about financial damages stemming from AI chatbot errors and hallucinations - occurrences where AI models generate false information with confidence.

The insurance policies are offered through the startup company, Armilla , and cover costs associated with legal claims against a business if a customer or third party is harmed by an underperforming AI product.

As companies increasingly deploy AI to enhance operational efficiency, some tools - like customer service chatbots - which have made costly and embarrassing mistakes.

Development & Operational Aspects

The development of these AI-related insurance products involves rigorous testing, validation, and extensive data collection, often with input from industry experts. The platforms leverage advanced algorithms combined with vast datasets, including property information, historical claims insights, and real-time market trends. This integration allows insurers to deliver precise cost estimates quickly, at a lower cost, and to better manage risks.

Risks & Accountability

Acting on hallucinated or false information can have serious consequences—leading to flawed decisions, financial losses, and reputational damage.

A pressing concern is accountability: if responsibility is shifted from humans to AI, who is liable for errors?

Kelwin Fernandes, CEO at specialist advisory firm NILG.AI says that removing human oversight or relying on AI for decision-making raises complex liability issues. Often, the blame is directed at the organisations behind the AI. For example:

Virgin Money issued an apology earlier this year after its chatbot criticised a customer for using the word “virgin.”

In 2024, Air Canada faced legal action after its chatbot fabricated a discount during a customer conversation.

According to a Financial Times report, Armilla’s policies would cover losses from such errors if the AI underperformance resulted in financial damage.

Industry Trends & Challenges

Lloyds Bank has also been exploring AI adoption amid concerns about hallucinations, initially focusing on back-office processes with human oversight. Despite technological advancements, recent assessments by OpenAI suggest that newer AI models hallucinate more often than earlier versions.

The integration of AI into the insurance sector has brought about significant benefits: improved data processing, increased analysis efficiency, and enhanced predictive capabilities. AI-driven innovation fosters greater accuracy, agility, and problem-solving capacity, transforming the industry. However, challenges remain:

A lack of expertise and technical familiarity across the industry.

Growing concerns over cybercrime, data security, and privacy risks.

The urgent need for cybersecurity professionals to mitigate evolving online threats.

Conclusion

While industry leaders like OpenAI and Google have made strides in reducing hallucination rates, the issue persists - especially with advanced reasoning models. The intersection of AI and insurance continues to evolve rapidly, promising increased efficiency and innovation, but also demanding careful management of risks and liabilities.

For sustained growth, the insurance industry must address the talent gap in cybersecurity and data management while fostering trust in AI systems.

Image: Ideogram

You Might Also Read:

Traditional Cyber Insurance Isn’t Built For AI-Driven Attacks:

If you like this website and use the comprehensive 8,000-plus service supplier Directory, you can get unrestricted access, including the exclusive in-depth Directors Report series, by signing up for a Premium Subscription.