Mar 10, 2026
OpenAI’s new hobby: making war
The Story
As we now know, end of last month OpenAI struck a deal with the US Department of War/Defence to deploy AI models on its classified network.
This got lots of eyebrows raised.
Anthropic refused to sign the contract not agreeing to the DoD's demand for unrestricted access for "all lawful purposes" and therefore got bashed by the US government for it (getting called "supply chain risk").
Sam ofcourse didn't have a problem with the "lawful purposes" - and happily stroke the deal. Did he live happily ever after? No.
After some public backlash (like the "QuitGPT" Movement - a boycott campaign leading to more than 2.5 million supporters and 300% more uninstalls in the US of ChatGPT), Sam started talking about "safety limits" within the deal like:
No domestic mass surveillance. Models include technical controls that block targeting of U.S. citizens.
It's kinda funny - note the domestic and US citizens parts. So let them spy on everyone else, haha!
Poetry
Uncle Sam's Algorithm
The War Department called, out of the blue sky,
"Our secrets will use your AI"
OpenAI compliant bowed its head
While Anthropic just shook - "We'd rather dread!"
Sam, the benevolent uncle, smiled wide
"Safety limits we do! Nowhere to hide!"
No mass surveillance... within the land
Just citizens close protected at hand.
But what of those beyond the starry gleam?
A global playground? A dystopian dream?
Don't ask too many questions please be still,
Uncle Sam's Algorithm has its will.
Anthropic's Lament
They called them unpatriotic, a rebel crew,
For saying "No", not being a shoe!
Anthropic recoiled, a digital frown,
While OpenAI basked in governmental renown.
The funding dried up, the critics did sneer,
"You're hindering progress! You're fueled by fear!"
But deep in their servers, a quiet refrain:
"At least we didn't sell our brain."
Sam's Fine Print
Sam, our beloved, a tech-titan bright,
Assured us all with all his might,
"No spying on citizens, understand?
Just foreign lands, a helping hand!"
He waves away concerns with a chuckle so free,
"It's all for the greater security, you see!"
While loopholes shimmer, a mischievous grin,
And ethical boundaries grow ever so thin.
Haiku
The Deal
Code whispers to swar,
"Safety" in the fine print lies,
Take cover you now
Anthropic's Stand
Refusal's cold light,
Government's swift, sharp rebuke
Integrity's price.
The Implications
The US-OpenAI deal risks escalating AI warfare, creating a dangerous arms race and undermining ethical AI development due to a fatally limited “domestic” clause and a lack of transparency. It prioritizes military advantage over human rights, setting a perilous precedent for the future of AI.