Loss run

In the AI space, Bavaya is setting a new benchmark with its specialized InsurGPT trademark model built specifically for the insurance sector.

The model achieved an impressive 93.1% field accuracy on loss runs well ahead of general purpose AI models, which scored between 78% and 85%.

Bavaya's results highlight the value of targeted training and show how sector-specific development can streamline submissions and claims workflows while reducing the need for manual review.

Back on the physical risk front, the Insurance Institute for Business and Home Safety has released new findings from its IBHS Roof Materials Aging Farm program.