Increasing employee diversity can benefit not just the public perception of an insurer and the pride of its workforce—it can also help expand its customer base. By creating a safe space and sense of belonging for diverse employees to express their views, companies can create innovative products and solutions that can meet the needs of underserved markets.
This was just one of the insights shared at Verisk Velocity 2020 at a session entitled “Combating Racism with Ethical Artificial Intelligence.” The session, which was led by Verisk’s vice president of inclusion and diversity, featured a compelling discussion on race, insurance, and related challenges posed by artificial intelligence (AI). The discussion featured an experienced leadership coach, a state insurance regulator, and Verisk’s chief analytics officer.
Below are three insights from the session:
Increasing diversity requires small steps, consistently over time
Building a diverse workforce requires a range of efforts, some of which are easier to achieve than others. Increasing the diversity of the c-suite and creating an inclusive work culture can make a big difference but can take significant time and dedication. Developing internship programs for minorities, providing diversity training, and translating insurance materials into other languages are smaller efforts but can still help attract and retain a demographic that is closer to representing society at large.
AI is not the problem. It’s how we develop it.
AI often takes the blame in analytic decision-making, but it's humans who need to take responsibility and look deeper into the data. For instance, if criminal history is considered in underwriting, black applicants may face discrimination in situations where they’re more likely to be arrested than white people due to racial bias in the criminal justice system. Insurers need to look carefully at how the analytics inform their decisions and affect both existing and potential customers. Then, they need to retrain the AI to make more ethical decisions that fight bias in data.
AI needs to be fair, safe, and transparent.
To serve the diverse marketplace, insurers need to strive for AI that is fair, safe, and transparent. That transparency is critical, so customers can understand how insurers arrive at their underwriting decisions. The ethical underpinning is also important, because many decisions will be hard to navigate. When insurers take into consideration the needs of underserved groups, they can build a trust with their customers that can help their business grow and be profitable.
Learn more about these themes and others by watching the session on-demand at Verisk Velocity.