Ghosts In The Machine

How AI Is Transforming Your Industry

Once slow to adapt, now advancing rapidly

by John Thornton

Mr. Thornton is Executive Vice President, Sales & Marketing with Amalgamated Life Insurance Company. Visit

While the insurance industry has been slower than other industries to adopt new technologies, it is collectively getting behind Artificial Intelligence (AI) and related technologies like Machine Learning (ML) to improve operations. AI is transforming processes ranging from risk assessment and underwriting, fraud detection, and risk mitigation, to claims processing and customer service.

The global AI in insurance market size was estimated at $4.59 billion in 2022 and is projected to reach $79.86 billion by 2032 with a compound annual growth rate (CAGR) of 33.96% from 2023 to 2032. (Source: Precedence Research). By applying AI, insurance carriers and advisors are gaining valuable benefits enabling their resiliency and ability to remain competitive in a marketplace besieged by rising interest rates, geopolitical disturbances, and increasing regulation. AI, however, is not without risks. Therefore, it is important that insurance companies understand this powerful and disruptive technology; its applications, benefits, risks, regulatory developments, and the need for its responsible use.

AI In Today’s Insurance Landscape

Forbes recently reported that the insurance sector experienced a 60% increase in operational efficiency by applying AI. The areas achieving the highest increases were claims accuracy showing up to a 99% improvement and customer experience improving up to 95%. Some North American insurance companies are operating the technology on-premises, while others are deploying it through the cloud. Offering a more flexible, scalable environment, cloud platforms facilitate pre-built AI tools and ML frameworks, as well as Application Programming Interfaces (APIs). This enables insurers to easily develop and integrate AI solutions into their existing systems on a pay-as-you-go frequency. Using cloud solutions, insurers can securely manage and access both structured and unstructured data which is essential for AI-driven analytics and informed decision-making.

There are several developments occurring in the AI-driven insurance sector. According to the National Association of Insurance Commissioners (NAIC), the industry’s expanded use of AI-enabled chatbots has increased the delivery of more personalized customer interactions, advice, and product marketing. When multifunctional chatbots are used, insurance companies can promptly respond to customer inquiries, support claims processing, make policy recommendations and detect fraud. The application of the Zuri AI chatbot was reported to have resolved 70% of insurance queries without any human communication. (Source: Markets and Markets). Advanced AI chatbots have the ability to understand complex inquiries, process large amounts of data, and then provide real-time, actionable, insights.

AI is adroitly helping to address some of the insurance industry’s greatest challenges. Most notably, as insurance companies continue their process digitalization, gaining greater speed, efficiency, and accuracy, they become even more vulnerable to fraud and other online cybercrimes. As one of the most data-intensive industries, the insurance sector must not only manage and leverage its data, but it must do so while ensuring data privacy and security. Data breaches in the industry come with hefty financial penalties, not to mention the loss of customer trust and reputational damage. Building a sound strategy related to data protection when applying AI and ML, which transforms data into knowledge and action within the context of sophisticated AI models (e.g., chatbots, robo-advisors), can be challenging. Further, many insurance companies are struggling to find skilled AI specialists to provide support in these areas. This also leads to another related challenge – the potential for an over dependency on digitalized, automated processes without the appropriate human oversight resulting in errors or lost opportunities. The need is for insurance companies to secure highly qualified staff and develop AI training programs to cultivate upskilling in existing staff.

Also, not to be overlooked while embarking on AI initiatives is to develop strong legal and governance procedures to ensure regulatory compliance. In September 2023, the Geneva Association, the only global association of insurance companies, whose members manage $21 trillion in assets, released a report highlighting the “evolving AI regulatory landscape for insurers.” The report analyzes various approaches to AI regulation and how these applications impact on the industry and support innovation, while ensuring adequate protection for customers by leveraging existing “technology-neutral insurance regulatory frameworks to manage AI-related risks.”

NAIC’s 5 AI Principles

In the United States, the NAIC set forth five AI principles for insurance companies to follow which are:

  • Fair and Ethical – AI actors (users) should proactively engage in responsible stewardship of trustworthy AI in pursuit of beneficial outcomes for consumers and to avoid proxy discrimination against protected classes.
  • Accountable – AI actors should be accountable for ensuring that AI systems operate in compliance with these principles.
  • Compliant – AI actors must have the knowledge and resources in place to comply with all applicable insurance laws and regulations. Compliance is required whether violations would be intentional or unintentional.
  • Transparent – AI actors must have the ability to protect confidentiality of proprietary algorithms and demonstrate adherence to individual state laws and regulations in all states where AI is deployed.
  • Secure, Safe and Robust – AI systems should be robust, secure, and safe throughout the entire life cycle so that in conditions of normal or reasonably foreseeable use, they can function in compliance with applicable laws and regulations.

Regulatory Climate

Recently, President Biden released The AI Executive Order mandating many government agencies to develop areas of AI regulation in 2024. It also encourages open development of AI technology innovations including those focused on new AI security tools. Covered in the Executive Order are topics such as: how AI-generated code can be exploited and what measures organizations can take to reduce their risks, the creation of a registry of AI models, large foundational AI models, AI safety, civil liberties and algorithm bias, AI-generated code, and cybersecurity.

As insurance companies continue their process digitization, gaining greater speed, efficiency, and accuracy, they become even more vulnerable to fraud and other online cyber-crimes...

At the state level, several states have been developing laws to govern AI’s application in insurance. Colorado developed the first state enacted legislation in July 2021, which went into effect in 2023, titled Restrict Insurers’ Use of External Consumer Data. Its requirements extend beyond the usual unfair discrimination standard to require insurers’ reporting external data sources used to develop and implement algorithms and predictive models; explaining how external data sources are being used; establishing and maintaining a risk management framework and assessing its results; acting to minimize risks of unfair discrimination; and attesting that the risk framework was implemented appropriately and on a continuous basis by the insurers’ Chief Risk Officer.

California’s Insurance Commissioners issued an order requiring insurance companies and other licensees to avoid both conscious and unconscious bias or discrimination often resulting from the use of AI and other Big Data models. Other states which have issued various laws/regulations relating to AI’s application include Connecticut, Louisiana, and New York, as well as the District of Columbia. As regulatory developments continue, insurance companies remain committed to leveraging AI to improve their operations.

AI Applications in Insurance

AI has proven effective in improving various insurance processes as follows:

  • Underwriting – AI has been instrumental in leveraging its ability to capture data to automate certain underwriting processes, by streamlining information harvesting and document review, thereby enabling underwriters to focus on high-value and more complex tasks. It also supports a rules engine that assists underwriters when rules are automatically applied for claims adjudication without requiring human involvement and real-time binding of contracts. In underwriting, AI also can help make risk assessment more precise and enable more accurate premium predictions based on past risk assessments.
  • Claims Processing – Without human intervention, AI can handle the initial notice of loss, supporting insurers’ reporting, routing, triaging, and assignment of claims. AI-chatbots can facilitate claims reporting whereby customers can report their situation from any device, anywhere, anytime and the AI-enabled chatbot can efficiently and expeditiously distribute their information for additional processing.
  • Claims Management – AI and ML can be used with other applications and can streamline the claims management and investigation process through their regulation of all processes and efficient data/information capture, authorizations, approvals, payment tracking.
  • Appeals Management – Applying AI, along with Optical Character Recognition (OCR) and Robotic Processing Automation (RPA), insurers can realize a faster, lower-cost appeals process, while also enhancing the customer’s experience.
  • Financial – AI can enable automated document generation, payments, and invoices. Additionally, it can support informed market forecasting through its evaluation of financial news reports and can generate investor financial reports based on structured input, while also detecting potential opportunities or challenges based on its assessments.
  • Data Harvesting – AI has enabled insurers’ collection of data from various sources and data services to access real-time data relating to finances, litigation history, business practices, etc.
  • Process Automation – AI drives robotic process automation of repetitive or redundant tasks.
  • Risk Assessment – AI, in conjunction with ML, can flag potentially fraudulent claims, supports anti-money laundering (AML) compliance through its ability to detect fraud by analyzing historical data, identifying suspicious patterns and anomalies, anticipating risks and identifying emerging risk trends, and preventing fraud, as well as reducing human error.
  • Personalized Customer Service – Through its use of data, AI helps insurers develop a better understanding of their customers, enabling them to develop customized, personalized marketing materials and sales messages. This, in turn, helps increase customer engagement, conversion rates, customer retention, cross-sell opportunities, and revenue. AI can further support an enhanced customer experience by enabling the integration of virtual agents into a customer service platform.

Responsible AI

In a call for trustworthy AI, INFORM, a global pioneer in AI-driven optimization software, headquartered in Aachen, Germany, recently released its Responsible AI Guidelines for all organizations across diverse industries to follow.

The guidelines revolve around several pivotal principles:

  • Beneficial AI – Ensuring AI systems enrich both users and society and mitigate negative impacts like bias and misinformation.
  • Human-centric AI – Promoting AI’s supportive role to humans, enhancing decision-making processes, and upholding human responsibility.
  • Aligned AI – Guaranteeing AI is in sync with human and business values, with clear and understandable AI as a foundation.
  • Privacy-preserving AI – Upholding European Union’s GDPR standards and achieving top-tier security standards endorsed by ISO 27001 certifications.
  • Reliable AI – Prioritizing quality, consistency, and transparency in AI applications, especially in vital sectors.
  • Safe AI – Crafting AI algorithms that ensure safety, and ward off potential threats.

The insurance industry’s recognition that an enterprise-wide strategy for strong AI governance and risk management is important, but not solely sufficient. Responsible guidelines that align with business strategies, compliance, and ethical use of AI form the critical cornerstone.