There you have it! Emil Michael handed Google a near-monopoly on the DoW's entire AI market — and…
May 4, 2026 · 0 likes · 0 comments
Cybersecurity AI
There you have it! Emil Michael handed Google a near-monopoly on the DoW's entire AI market — and it has been nine months with zero course correction.
GenAI.mil. The DoW's flagship AI platform. Built to give warfighters cutting-edge AI tools across every branch, every mission set. Sounds great on paper.
The reality? Roughly 95% of GenAI.mil runs on Google. One vendor. One cloud. One model family. For the most consequential technology the DoW has deployed in decades.
That is not strategy. That is dependency.
I have been warning about this for years. The moment you tie your entire AI capability to a single provider, you are not building resilience — you are building a single point of failure at scale. And you are giving that provider leverage over every future negotiation, every pricing conversation, every architectural decision for the foreseeable future.
Abstraction is not a buzzword. It is the only sane approach to government AI deployment. Model-agnostic layers. Vendor-agnostic platforms. The ability to swap providers when performance shifts, when pricing changes, when geopolitics demand it. That capability is not optional — it is foundational.
Emil Michael's DoW has done the opposite. They signed massive agreements with Google, Microsoft, Amazon, Oracle, OpenAI, Nvidia, SpaceX, and Reflection AI. They cut out Anthropic entirely — calling Dario Amodei an "ideological lunatic" during congressional testimony.
So the DoW is picking winners and losers in the AI market based on politics, not performance. And the biggest winner — by a staggering margin — is Google, sitting at roughly 95% of GenAI.mil's footprint for nine months running.
Ask yourself: what happens when Google raises prices? What happens when a better model drops from a competitor and the DoW cannot pivot because the entire stack is welded to one vendor? What happens when Google's priorities diverge from national security needs?
The answer is simple. You are stuck. The DoW is stuck. The warfighter is stuck.
Abstraction was always the answer. Use the layer that lets you plug in any model, any cloud, any provider — and switch when you need to. I have been saying this day 1. The government needs vendor-agnostic AI infrastructure, not loyalty programs.
Nine months of a Google near-monopoly on the DoW's AI. And counting.
I have seen this movie before. It does not end well.
You have been warned.
GenAI.mil. The DoW's flagship AI platform. Built to give warfighters cutting-edge AI tools across every branch, every mission set. Sounds great on paper.
The reality? Roughly 95% of GenAI.mil runs on Google. One vendor. One cloud. One model family. For the most consequential technology the DoW has deployed in decades.
That is not strategy. That is dependency.
I have been warning about this for years. The moment you tie your entire AI capability to a single provider, you are not building resilience — you are building a single point of failure at scale. And you are giving that provider leverage over every future negotiation, every pricing conversation, every architectural decision for the foreseeable future.
Abstraction is not a buzzword. It is the only sane approach to government AI deployment. Model-agnostic layers. Vendor-agnostic platforms. The ability to swap providers when performance shifts, when pricing changes, when geopolitics demand it. That capability is not optional — it is foundational.
Emil Michael's DoW has done the opposite. They signed massive agreements with Google, Microsoft, Amazon, Oracle, OpenAI, Nvidia, SpaceX, and Reflection AI. They cut out Anthropic entirely — calling Dario Amodei an "ideological lunatic" during congressional testimony.
So the DoW is picking winners and losers in the AI market based on politics, not performance. And the biggest winner — by a staggering margin — is Google, sitting at roughly 95% of GenAI.mil's footprint for nine months running.
Ask yourself: what happens when Google raises prices? What happens when a better model drops from a competitor and the DoW cannot pivot because the entire stack is welded to one vendor? What happens when Google's priorities diverge from national security needs?
The answer is simple. You are stuck. The DoW is stuck. The warfighter is stuck.
Abstraction was always the answer. Use the layer that lets you plug in any model, any cloud, any provider — and switch when you need to. I have been saying this day 1. The government needs vendor-agnostic AI infrastructure, not loyalty programs.
Nine months of a Google near-monopoly on the DoW's AI. And counting.
I have seen this movie before. It does not end well.
You have been warned.