About a couple of years ago, when AI turned again into a hot topic in the tech world, there was a lot of enthusiasm for what it could do. But also fear for possible jobs losses. Nowadays the enthusiasm has been curbed due to the bizarre experiments done so far: AI should only be used to speed up simpler tasks. But a human should always make the final decision. In the case of the insurance industry, brokers and underwriters would turn into cyborgs.
This and other themes were discussed at Insurtech London AI & Algorithms, which happened last Monday (03) at The Steel Yard.
The insurance industry has been using statistical models for years. But with the popularity of the term AI, some decided to join the hype and market it as such. That resulted in a lot of talks and no show and lack of trust in what is or not AI. Even if it is and it is learning and changing over time better than us, how can we trust it? There are at least a couple of cases that can easily be found online of AIs that went awry and started doing things that weren’t expected and scientists couldn’t tell why. That challenges insurers’ due diligence.
One solution to this suggested by Peter McBurney, Head of Technology Consulting at Norton Rose Fulbright, is applying Explainable AI (XAI). As the name suggests, it’s AI that can explain their results in a way a human could understand. As opposed to the concept of “black box” in Machine Learning where even their designers cannot explain its results.
Regardless of which method is applied, gathering quality data is a challenge. People are increasingly conscious of their own data and all these technologies are fed by it. Where and how should insurers collect it in an ethical way? Some are using publicly available data and others partnering with other companies that have the data they need.
Finally, the last concern that was raised by some of the speakers was the fact that the insurance industry is not doing enough to attract the best tech talent. Insurers need to rebrand themselves and show the positive impact it has in society if it wants to be the first choice of the new AI tech talent.