One of the biggest questions is whether

companies should be allowed to create

and enforce their own rules around AI

independently of government regulation.

We have seen a similar event in Iron Man

2 where Senator Stern asks Tony Stark to

turn over his Iron Man suit to the US

government. And he responds with, "I've

successfully privatized world peace.

What more do you want?" Similarly,

Anthropic has their own powerful

technology that the US government is

asking for full military purposes and

they want unfathered access to Claude,

which means Anthropic needs to walk back

on the very thing that they built their

business around, which is AI safety.

Now, Anthropic has their own flaws in

their business model, but they are the

closest thing to AI safety that we have

compared to other AI frontier labs. And

the recent power struggle between

Department of Defense Secretary Pete

Hacksth and Anthropic CEO Dario Amade

sets a huge precedent on how exactly the

relationship between private companies

and the US government should look like.

Welcome to Caleb Bright's Code where

every second counts. Quick shout out to

Zo. More on them later. Now, out of the

15 executive branches, the Department of

Defense has the biggest in human capital

and they're the biggest employer in the

US. And it's this department that

granted $200 million to other four

frontier labs in the US back in July

2025, totaling up to $800 million

combined. But Anthropic has been working

with the US government outside of the

Department of Defense since 2024, where

after they released Sonnet 3 for the

first time, they provisioned Claw 3,

Haiku, and Sonnet for the government

through AWS dedicated cloud called

GovCloud. Enthropic also partnered up

with Palunteer in November that year to

provide the US intelligence and defense

agency with a system that can help

deploy CLA to process information

faster. And in August 2025, they also

gave one-year access to all branches of

the US government to help AI adoption

across the government. As you can see,

Enthropic had already rooted themselves

deep into the government system by

providing claude through Palunteer

GovCloud with AWS and more. Now, fast

forward to today. Even though four other

Frontier Labs also got the same $200

million deal from Department of Defense

back in July 2025, Enthropic has

specifically been given an ultimatum by

Pete Hackath on February 24th to

essentially drop their safety guard rail

on the model. And the consequences for

not following this would result in three

potential outcomes. First, be labeled as

a supply chain risk, which would cut all

military contracts with anthropic.

Second, be forced to compel by enacting

the Defense Production Act, which is a

draconian measure. Or third, terminate

the $200 million contract that were

given in the first place. Now, you might

have noticed by now that the size of the

contract is only 1% of Enthropic's

entire revenue for the year. So, this is

already more about the principle than

about the money per se. While in

principle, Enthropic and the Department

of Defense both would agree that there

are lines we shouldn't cross when it

comes to how AI should be used in such

cases like mass surveillance and fully

autonomous weapons. This is more about

who gets to make the final authority to

draw those lines. And depending on the

polling, American people also have

differing views on this very question.

Whether private companies should have

the ability to set their own ethical

rules around their product and what

finally happened between Anthropic and

Department of Defense is quite crazy.

And here's a quick word from Zo.

OpenCloud showed the world what a

personal AI agent could look like. But

if you actually try to set it up, it's

actually not as easy to set it up

properly. Zo takes a different approach

by turning your laptop into a server. Zo

gives you your own computer in the cloud

with AI builtin. It's fully isolated, so

you're not one bad configuration away

from exposing your whole machine. Out of

the box, you can integrate with Gmail,

Google Drive, Stripe X, Spotify, and

more. You can also build websites, set

up automations that run while you sleep,

and even code full projects right from

chat. My favorite feature is texting Zo

on iMessage or Telegram and talking to

my agent. Zo is always on, always yours

without configs. Link in the description

below. After two long days of standoff

between Enthropic and the Department of

Defense, the official statement was made

by Peak Hacks to designate Enthropic as

a supply chain risk to the national

security. This was of course after

Daario made a statement standing his

ground on such matters. Now, one quick

clarification to be made here is that

the extent of the ban that applies to

Anthropic only applies to contracts

associated with the Department of War,

not the entire branch of the US

government. Enthropic was given 6 months

to transition their already deeply used

claude models out and what now looks

like open as model coming in. And

Enthropic's response to being labeled as

a supply chain risk was standing their

ground on their decision. Dario made a

statement how the US has never publicly

labeled an American company a supply

chain risk. He also commented that

frontier AI models just simply aren't

there yet to be used in more

sophisticated applications like fully

autonomous weapons. And the fact that

Enthropic treated AI safety as a core

part of its business likely contributed

to being an early adopted model by the

US government. And it's this very thing

that seem to help them win government

contracts in some but also lose in

others. But in all fairness from the

Department of Defense's perspective,

they claim that Anthropic is essentially

trying to strongarm the military by

using what they call effective altruism

to control how they think the military

should be using their product. And while

a reasonable response could have been to

just terminate the contract short of

coming off weak in negotiation, what

really won the public opinion is

anthropic, not the department of

defense. For now, Sam Alman announced

that OpenAI will be providing their

models to the Department of Defense, but

basically under the same conditions as

Anthropic made. And you might be

wondering why OpenAI got a different

outcome than Anthropic despite making

the same stance. OpenAI's agreement with

the military is in principle that yes,

AI shouldn't be used for such cases. But

on the other hand, anthropic stance is

much more concrete than pure principle

since safeguards in place are baked into

the model's weights than in principle.

At the end of the day, this entire

ordeal between Anthropic and the US

government sets precedent on how

government and Frontier Labs should work

together under the powerful technology

that's being built outside of much

government regulation. And given that

we're still really early in finding out

just how powerful AI really could

become, the tension between the military

use cases and private frontier labs will

be something that we'll need to continue

monitoring going forward. What do you

think? Do you think the Department of

Defense was well within their rights to

ban anthropic, but only to the extent of

the military contracts? What do you

think the government and private AI

companies should look like in the

future?