Blog: Ethical data – building trust & combatting bias

Data management strategies that work are built on trust. That means solid privacy policies and proper compliance. In the second of a three-part series, Vikas Bhalla, EVP and Head of Insurance at EXL, explores how ethics can be central to creating transformative data programs.

Put simply, data sharing is an exchange of value between an insurer and a customer. The customer shares his data and in return expects optimized products and services without his privacy being jeopardized. Without trust, the process collapses.

Putting consumers in control

From the impenetrably complicated rules of the General Data Protection Regulation to the new ones, consumer fears around data use are understandable United Kingdom Data reform bill pretending to back down EU Bureaucracy”. Little detail in such laws reaches consumers, and the world of regulation often seems shrouded in mystery to the ordinary citizen on the street.

Insurers must communicate to customers what data is collected, how they use it and how it is kept secure. That way, they feel empowered to make choices. Once trust is established, the benefits of participation are significant. Whether it’s telematics helping a driver prove they’re not liable in a car accident, or smart home devices reducing home insurance premiums by detecting everything from floods to burglaries.

Measure and manage privacy risks

Insurers must also take steps to assess and mitigate privacy risks. Taking these actions and making it clear to consumers what they are doing and why they are doing it is critical to building trust and establishing a reputation as a responsible, data-driven insurer.

For example, we partnered with a global insurer to quantify and mitigate risk in over 40 countries with the aim of ensuring compliance with regional data protection laws such as GDPR GDPR in which United Kingdom/EU and other applicable guidelines for the USSouth America and other areas.

Working with data experts on programs like this can unlock the mystery of data governance and give consumers the protection they deserve by keeping their private information private. From here, insurers need to be crystal clear on what steps they will take in the event of a data breach or incident, and communicate this effectively to their consumers, as well as how such an event would be remedied.

combating prejudice in AI

Finally, data usage must be free of bias. preload on AI has surged into consumer awareness in recent years through media headlines with stories of sexist recruitment bots and racist bank lending software.

Every industry uses AI need to be aware of the risks of bias and how to mitigate them. The cause of bias in algorithms often lies in their design or training. For example, if a machine learning algorithm to predict credit risk was only trained on high net worth individuals, most real customers would never qualify.

To prevent these biases from occurring, algorithm audits must be performed. When you do this on a regular basis — examining how data is analyzed and sorting out inconsistencies — insurers can rest assured that their algorithms are fair and accurate.

Ethical insurers must lead

With the big online retailers – and their wealth of consumer data now pouring into the insurance market – the ethical use of data is an issue that will increasingly find its way into boardrooms.

A consumer-centric strategy with a robust risk management and bias prevention approach will help insurers succeed. When consumers ask for their data, insurers must be prepared to defend their position as ethical, data-driven providers.