Insurance
Home›Insurance›Industry & Deals›Insurance AI pilots often stall after launch due to un…
Insurance AI pilots often stall after launch due to unclear ownership
Only 7% of insurers had brought AI systems to scale as of 2025, while 45% were still exploring AI, according to BCG.
Insurance AI projects are increasingly running into an operational bottleneck after pilots end, with carriers struggling to determine who owns the model once it moves into production, Insurance Business reports. As insurers shift from experimentation to real-world deployment, the post-pilot handoff has become a make-or-break question.
BCG reported in 2025 that only 7% of insurance companies had successfully brought AI systems to scale, despite the industry being well positioned for AI given its data reserves and analytics workforce. In parallel, a separate 2025 insurance AI adoption survey found 45% of respondents were still exploring AI, 25% were testing discrete use cases, and 22% were running live solutions in production.
According to Rob Galbraith, CEO of Forestview Insights and a former insurance innovation leader, the gap is not simply a vendor problem, it is a product-management issue within the carrier. He argues that turning technology into a product requires account ownership, billing and implementation support, an updates process, a path for patches, and defined procedures for when performance changes.
Galbraith also describes a common failure mode for internal builds, where projects close after launch and the specialists needed to maintain the model move on. “Once it implements, those resources are gone,” he said, adding that subject-matter experts can also end up without a clear role, leaving companies without the operating structure to keep the model useful.