Post

To strike 1,000 targets in 24 hours in Iran, the U.S. military leveraged the most advanced AI it’s ever used in warfare. Anthropic’s Claude partnered with the military’s Maven Smart System, suggesting targets and issuing precise location coordinates. https://wapo.st/46BFe1T

223 reposts480 likes

So you are now using the Psychosis Hallucination Machine to decide where to lob missiles. I guess that explains the incinerated school and all the dead children.

I guess the techbros didn’t think ruining the kids’ future environment on behalf of AI was evil enough, so they just ended their present.

It’s great that WaPo is writing puff pieces for the U.S. military, when one of the first things that happened was the murder of 150+ children in a school Do you guys still have trouble sleeping at night, or was selling your soul away enough?

Another reason to say FUCK AI* * I can accept AI IF it can be used to speed up Medical and Scientific research but we don't need it in everything we do. The post above is just tone deaf about the human cost of what has just happened

You may be right. But, even for a fascist/authoritarian collaborator and propaganda outlet, extolling the virtues of technology probably responsible for numerous war crimes (schools, hospitals) is not good for subscriptions.

How did the Ai determine those kids were a military target. Seems like something we should look into. You know if I was an investigative reporter work at a large newspaper that’s something I’d certainly do.

What a mess. For what it's worth, it looks like Rubio let the truth out when he said they came in because they knew Israel was launching an attack. The rest of it, all their statements and sloppy attempts to justify are a pack of lies and total rubbish.

This was OpenAI, not Anthropic. Pentagon Anthropic deadline was 5 PM EST Feb 27, but Trump ordered ban early, @3:47 PM EST on Truth Social. Order to hit Iran given Feb 27 @ 3:38 PM EST. Major operations started Feb 28 @ 1:15 AM EST, school hit @2:15 AM EST. Anthropic & Pentagon resumed talks Mar 5.

Wait until captain drunk pants has a bad day and sends it's all the negative social media posts about him and tells AI to assassinate everyone. So much for the GOP and government overreach.

When early strikes included a school and hospital, I'm not sure that's the brag anytime involved seems to think it is. More of an indictment, actually.

Did Anthropic's Claude write this glowing, sensual copy about how advanced and sexy Anthropic's Claude is? Also, are we bragging about demolishing an elementary girls school? Wtf WaPo. Like we know you are part of the machine but fake it a LITTLE, christ.

So the DoD declared Claude a public safety risk, announced all federal contractors and departments had to shut it down immediately on Friday night, and then went ahead and used it on sensitive and classified national security information?

This was OpenAI, not Anthropic. Pentagon Anthropic deadline was 5 PM EST Feb 27, but Trump ordered ban early, @3:47 PM EST on Truth Social. Order to hit Iran given Feb 27 @ 3:38 PM EST. Major operations started Feb 28 @ 1:15 AM EST, school hit @2:15 AM EST. Anthropic & Pentagon resumed talks Mar 5.

This was OpenAI, not Anthropic. Pentagon Anthropic deadline was 5 PM EST Feb 27, but Trump ordered ban early, @3:47 PM EST on Truth Social. Order to hit Iran given Feb 27 @ 3:38 PM EST. Major operations started Feb 28 @ 1:15 AM EST, school hit @2:15 AM EST. Anthropic & Pentagon resumed talks Mar 5.

They killed kids "Here lies the fundamental contradiction...If the intelligence was up to date enough to spare a clinic that had been open for only 1 year, how did it fail to identify an elementary school that...had [been a] civilian institution for [over] 10 years" www.aljazeera.com/news/2026/3/...

aljazeera.comAl Jazeera investigation: Iran girls’ school targeting likely ‘deliberate’Al Jazeera investigation raises questions over deadliest single attack of war that killed 165 schoolgirls and staff.

*A war crime committed with a computer, is simply a war crime* If the media is confused, the law is quite clear: bombing a grade-school with children inside is mass murder- all the way up the chain of command to the White House.

Didn't the US use half of its ammo stockpiles to kill 3k of Iranian soldiers, which accounts for 0.3% of their military? Plus, based on footage shared by the US army, a non-insignificant amount of ammunition was spent to bomb decoys painted on the ground. I'd mention civilians but you don't care.

Did you mean: Anthropic specifically refused to participate, they were put on a blacklist only used for enemies of the state, and yet their now-stolen algorithm was then used to murder school children?

This was OpenAI, not Anthropic. Pentagon Anthropic deadline was 5 PM EST Feb 27, but Trump ordered ban early, @3:47 PM EST on Truth Social. Order to hit Iran given Feb 27 @ 3:38 PM EST. Major operations started Feb 28 @ 1:15 AM EST, school hit @2:15 AM EST. Anthropic & Pentagon resumed talks Mar 5.

OpenAI was not integrated into current classified systems at the time of the strike. The article is correct that Claude was the code used for targeting.

Vengeance will be calculated and assured. The days of getting on a jet and flying somewhere carefree are over. Skyjackings. Cruise ships overtaken and commandeered by those seeking retribution for collateral damage in wartime. Reprisals will be sustained/brutal. We pay for their wickedness, not them

So it aimed for the school? Precisely? I say since you cant hold AI responsible for sloppiness, you just keep going up the line til you reach a person, and increase the fallout exponentially each level. They allowed this to happen. Precisely.

At this point does anything matter? You can't make this shit up anymore. It does it for ya

The Washington Post@washingtonpost.com

To strike 1,000 targets in 24 hours in Iran, the U.S. military leveraged the most advanced AI it’s ever used in warfare. Anthropic’s Claude partnered with the military’s Maven Smart System, suggesting targets and issuing precise location coordinates. https://wapo.st/46BFe1T

So they precisely targeted a school killing over 160 children? They precisely targeted nine hospitals? They precisely targeted paramedics helping people? They precisely targeted non-military cultural targets? That is what you are saying?

This was OpenAI, not Anthropic. Pentagon Anthropic deadline was 5 PM EST Feb 27, but Trump ordered ban early, @3:47 PM EST on Truth Social. Order to hit Iran given Feb 27 @ 3:38 PM EST. Major operations started Feb 28 @ 1:15 AM EST, school hit @2:15 AM EST. Anthropic & Pentagon resumed talks Mar 5.

OpenAI, not Anthropic. Trump banned Anthropic less than 60 minutes before the school was hit on Feb 28; negotiations didn't resume until March 6. OpenAI agreement with Pentagon was active Feb 27.

We are so beyond the point of “do you support this war or not”. The only acceptable position for the Democratic Party is “we are going to hold you accountable for war crimes”. “We are going to make up for letting Bush and Cheney walk”. Anything short of that is guilt by collaboration.

Young girls. Revolutionary guards. It's an easy thing to confuse. Early days, later models will be able to tell dogs from cockroaches. The latter will be useful as the board of Anthropic is apparently staffed with them.

How do you people live with yourself? Or is this AI? If its a human writing for the love of god get a job where you dont have to hate yourself

If the intended outcome was to burn small girls to death then yes that's very advanced technology that everyone should feel good about adopting into their daily life

This was OpenAI, not Anthropic. Pentagon Anthropic deadline was 5 PM EST Feb 27, but Trump ordered ban early, @3:47 PM EST on Truth Social. Order to hit Iran given Feb 27 @ 3:38 PM EST. Major operations started Feb 28 @ 1:15 AM EST, school hit @2:15 AM EST. Anthropic & Pentagon resumed talks Mar 5.

It almost sounds like the Washington Post is whitewashing and trying to legitimise an illegal invasion by a bunch of psychopathic felons that has killed scores of innocent civilians, most of them children. What a fucking surprise.

Within weeks of giving the US military unguided and unlimited use of Anthropoc’s Claude AI, it proves how insane that is by hallucinating outdated information from ten year old maps to target a converted school building and annihilate 175 young school girls. The US Army IS Skynet. 😱

This was OpenAI, not Anthropic. Pentagon Anthropic deadline was 5 PM EST Feb 27, but Trump ordered ban early, @3:47 PM EST on Truth Social. Order to hit Iran given Feb 27 @ 3:38 PM EST. Major operations started Feb 28 @ 1:15 AM EST, school hit @2:15 AM EST. Anthropic & Pentagon resumed talks Mar 5.