There you have it! At least two Google DeepMind researchers have quit over Google's April deal le…
October 23, 2026 · 0 likes · 0 comments
China Threat Defense Cybersecurity AI
There you have it! At least two Google DeepMind researchers have quit over Google's April deal letting the Pentagon use its AI models in classified settings for "any lawful government purpose."
The BBC published its investigation on Oct. 21. One researcher said the deal makes her "feel ashamed to work at Google." The other says Google has no veto over how the Pentagon uses the models.
Here's what the contract says. It permits "any lawful government purpose." It also says the technology should not be used for domestic mass surveillance, or for autonomous weapons without appropriate human oversight.
Google isn't alone. OpenAI, Amazon, Microsoft, SpaceX, Oracle, Nvidia and Reflection have signed comparable agreements, and xAI has a Pentagon contract too.
Google says it reviews every contract against its AI principles, and that providing its commercial models on industry-standard terms "represents a responsible approach to supporting national security."
We have seen this movie before. In 2018, thousands of Google employees protested Project Maven, the Pentagon's drone-imagery AI program. That June, Google said it would not renew the contract.
That same year, The Intercept exposed Dragonfly, a censored search prototype Google was building for China. A few months earlier, in December 2017, Google had announced an AI research center in Beijing.
I'm glad that this time the sheep at Google leadership didn't fold.
In 2018 they let a short-sighted, clueless crowd of employees win. That crowd was fine with a company that kept building for China, but walked away from the Pentagon. Maven got dropped. Our warfighters paid for it.
Not this time. Leadership held.
So to the folks walking out the door over this: good riddance.
Hopefully you end up far away from frontier AI, so your short-sighted, communist-friendly biases stay out of the models.
China isn't writing farewell letters about its AI. It's fielding it. The UBH piece even quotes Pete Hegseth: the United States will become an "AI-first warfighting force."
That's the right call. Google finally made it too.
The full story, just the facts: https://lnkd.in/ez6_d5Z3
The BBC published its investigation on Oct. 21. One researcher said the deal makes her "feel ashamed to work at Google." The other says Google has no veto over how the Pentagon uses the models.
Here's what the contract says. It permits "any lawful government purpose." It also says the technology should not be used for domestic mass surveillance, or for autonomous weapons without appropriate human oversight.
Google isn't alone. OpenAI, Amazon, Microsoft, SpaceX, Oracle, Nvidia and Reflection have signed comparable agreements, and xAI has a Pentagon contract too.
Google says it reviews every contract against its AI principles, and that providing its commercial models on industry-standard terms "represents a responsible approach to supporting national security."
We have seen this movie before. In 2018, thousands of Google employees protested Project Maven, the Pentagon's drone-imagery AI program. That June, Google said it would not renew the contract.
That same year, The Intercept exposed Dragonfly, a censored search prototype Google was building for China. A few months earlier, in December 2017, Google had announced an AI research center in Beijing.
I'm glad that this time the sheep at Google leadership didn't fold.
In 2018 they let a short-sighted, clueless crowd of employees win. That crowd was fine with a company that kept building for China, but walked away from the Pentagon. Maven got dropped. Our warfighters paid for it.
Not this time. Leadership held.
So to the folks walking out the door over this: good riddance.
Hopefully you end up far away from frontier AI, so your short-sighted, communist-friendly biases stay out of the models.
China isn't writing farewell letters about its AI. It's fielding it. The UBH piece even quotes Pete Hegseth: the United States will become an "AI-first warfighting force."
That's the right call. Google finally made it too.
The full story, just the facts: https://lnkd.in/ez6_d5Z3