The Chartered Insurance Institute is calling for insurers to tackle gender bias within their artificial intelligence (AI) algorithms.
Many insurers now use this AI technology within their claims, underwriting and marketing processes.
However a new report found that the datasets used to train the algorithms, which support AI system are “rooted in outdated gender concepts”.
The report points out that these systems are being ‘trained’ on historical data where decisions where based on a number of historic biases, particularly around gender.
It notes that more and more of the data is now unstructured, coming from text, audio video and sensors – but these historic biases will continue to persist unless properly addressed.
The CII says that insurance companies need to prepare a structured response to this issue, starting with visible leadership on tackling gender bias in AI. It also called for tools for testing historic data and algorithms for discriminatory features to be adopted, and training provided.
CII director of policy at public affairs at the CII says: “Tackling gender bias is both a fundamentally right thing to do, as well as a sensible, trust building move.
“The long-term winners of this insurance transformation will be those who consumers trust. Tackling bias should be woven into every insurer’s trust agenda.”
