Focus on EIOPA's report on ethical and governance principles for the use of AI in the insurance industry

Focus on EIOPA’s report on ethical and governance principles for the use of AI in the insurance industry

5 avril 2022

EIOPA, the European regulator for insurers, aims to foster financial stability and confidence in the insurance and pensions markets. This EU agency published a report in June 2021 on ethical and governance principles regarding the use of artificial intelligence in the insurance industry.

The integration of AI in the insurance industry can transform processes that are usually tedious and time-consuming: underwriting, claims management, fraud detection, customer service…
Thus, AI allows the implementation of:

– 24/7 conversational chatbots to respond to online requests;
– Automated email analysis and authentication systems for faster processing (preparation of response elements to be reviewed by the customer advisor);
– Rapid automated management solutions for small auto claims, with AI processing of the image of the wrecked vehicle;
– Analysis of data from connected objects (fall detector, water leak detector…) to trigger a human intervention.

The advantages of AI seem obvious for insurance players but an ethical and responsible use of AI is necessary. Many French and European rules already set a framework and limits to the use of data by insurance companies (RGPD, CNIL…). In April 2021, the European Commission unveiled its draft regulation on the ethical use of AI and is expected to legislate on AI liability in 2022.

The European Insurance and Occupational Pensions Authority, EIOPA, published a report for ethical and trustworthy artificial intelligence in the European insurance sector on June 18, 2021.

EIOPA’s expert advisory group report on digital ethics in insurance

EIOPA convened an expert advisory group that identified the opportunities and risks associated with the growing use of AI in insurance.
The experts then identified six key principles:

Proportionality Principle:
– Insurance companies should conduct an AI use case impact assessment to determine the governance measures required for a specific AI use case.
– The governance measures should be proportionate to the potential impact
of a specific AI use case on consumers and/or insurance companies.
– Insurance companies should then evaluate the combination of measures put in place to ensure ethical and trustworthy use.
Principle of fairness and non-discrimination:
– Companies must adhere to the principles of fairness and non-discrimination when using AI. They should consider financial inclusion issues and avoid reinforcing existing inequalities, especially for socially beneficial products.
Principle of transparency and explainability:
– Insurance companies should strive to use explainable AI models and tailor the types of explanations to the use cases and beneficiary stakeholders.
Human oversight principle:
– Insurance companies should clearly assign and document the role
and responsibilities of personnel involved in AI processes embedded in their governance system.
– It is also important that insurance companies assess the impact of AI on the work of employees and that employees have received adequate training.
Recordkeeping Data Governance Principle:
– Provisions included in national and European data protection laws (e.g., GDPR) should serve as a basis for implementing good governance of data used by insurance companies.
They must ensure that it is accurate, complete and appropriate whether it comes from internal or external sources.
– In addition, data should be stored in a safe and secure environment and, particularly for high-impact use cases, insurance companies should maintain appropriate records of data management processes and modeling methodologies to enable traceability and auditability.
Robustness and Performance Principle:
– Insurance companies should use robust AI systems, both when developed in-house or outsourced to third parties, taking into account
their intended use and potential to cause harm.
– AI systems should be deployed in resilient and secure IT infrastructures, including against cyber attacks.- Insurers, unanimous on the need to implement AI that is reasoned and respectful of the ethical principles they share, can implement these principles within their companies.