SBS Insurance Services has successfully concluded a pilot project with UnlikelyAI, trialling an advanced AI-powered platform designed to streamline insurance claims processing.
The pilot achieved automation of 40% of claims with a reported accuracy rate of 99%, while maintaining a comprehensive audit trail for each decision, the companies announced. The project marks the first public deployment of UnlikelyAI’s platform after several years of development.
UnlikelyAI’s system employs a unique combination of neural networks and symbolic reasoning, offering an alternative to traditional large language models (LLMs). Unlike LLMs—often criticised for inconsistent outputs and lack of decision transparency in regulated industries—the UnlikelyAI platform is built on rule-based logic, enabling consistent, explainable outcomes. This feature was a key factor in SBS Insurance Services’ decision to select the platform.
Throughout the pilot, the system handled claims with minimal manual input, routing only the most complex cases to human staff. SBS reported a 99% reduction in miscategorisation rates and noted significant improvements in processing efficiency. Crucially, every AI-driven decision was fully traceable, ensuring compliance with regulatory standards.
This collaboration reflects a broader trend within the insurance industry, as firms increasingly seek technological solutions that enhance operational performance while upholding strict compliance and customer service standards.
