As the U.S.-Israeli war on Iran continues, we look at how the Pentagon is using artificial intelligence in its operations. The system, known as Project Maven, relies on technology by Palantir and also incorporates the AI model Claude built by Anthropic. Israel has used similar AI targeting programs in Iran, as well as in Gaza and Lebanon.
Craig Jones, an expert on modern warfare, says AI technology is helping militaries speed up the “kill chain,” the process of identifying, approving and striking targets. “You’re reducing a massive human workload of tens of thousands of hours into seconds and minutes. You’re reducing workflows, and you’re automating human-made targeting decisions in ways which open up all kinds of problematic legal, ethical and political questions,” says Jones.
21 Comments
SpruceBringsteen@lemmy.world · 52 pts · 155d
Later, we'll find out the whole identifying part is a bit glossed over, and it's really about using the AI as a rubber stamping plausible deniability shield to commit whatever atrocity you feel like.
geneva_convenience@lemmy.ml · 18 pts · 155d
You mean oil and gas fields aren't Hamas strongholds? Well too late already bombed them.
IAmNorRealTakeYourMeds@lemmy.world · 48 pts · 155d
Israel did the same thing, by using AI, they can hallucinate as many targets as they can with the "plausible deniability".
Like the absolute worst plausible deniability in human history.
audaxdreik@pawb.social · 30 pts · 155d
There are just so many things to be said about the ills of AI, but one of them is that it is very purposefully a liability laundering machine. The decisions and thought process are blackboxed and unauditable. We've been trained to dismiss any oopsies as an inevitable part of the system, both while it's still "rapidly developing" as well as just inherent to the technology. Absolutely none of this is acceptable and yet here we are.
KurtVonnegut@mander.xyz · 4 pts · 155d
But how does it change liability? Isn't the person who decides to run this system ultimately responsible for its effects?
suicidaleggroll@lemmy.world · 12 pts · 155d
One would think so, but apparently not
themurphy@lemmy.ml · 3 pts · 155d
Everyone in the chain is reliable.
This is the same nazi argument. "I only transported the jews, I only watched them, I only build the concentration camps" and so on.
dx1@lemmy.ml · 1 pts · 154d
Yes.
SkavarSharraddas@gehirneimer.de · 11 pts · 155d
Implausible deniability.
baguettefish@discuss.tchncs.de · 5 pts · 155d
militaries slaughtering civilians using very shaky justifications that only work because they have the bigger fist or a strong backer probably isn't that unique, so you can't exactly use extreme words like "only", "worst", "ever" about current happenings. the IDF is only some of the worst scum possible.
IAmNorRealTakeYourMeds@lemmy.world · 2 pts · 155d
was referring to using BS AI to justify killing everyone. because AI that was designed to point at civilians said that civilians are targets.
it's the most obvious bullshit justification. that makes absolutely no sense. but he we are
GuyIncognito@lemmy.ca · 13 pts · 155d
After WW2, the industrialists who supported the Nazis mostly got off scot free. This was a terrible mistake and should not be repeated.
mub@lemmy.ml · 12 pts · 155d
Did I miss something. I thought Anthropic chose to not allow Claude to be used for military shit?
queermunist@lemmy.ml · 13 pts · 155d
They chose not to change their tools to comply with the new Pentagon demands. It's the Pentagon that then decided to cancel contracts and declare them a "supply chain risk" in retaliation. Claude is still integrated into their systems and it'll take time to switch.
eldavi@lemmy.ml · 4 pts · 155d
they did it too late and i bet that wasn't an accident.
UnimportantHuman@lemmy.ml · 3 pts · 155d
You think it was for show? I am genuinely curious of someone else's view because I've been skeptical myself. The fact they were even working with the Pentagon in the first place was a red flag to me.
eldavi@lemmy.ml · 4 pts · 155d
they had to be deeply involved with the pentagon to enable the anthropic ai the ability to function, so they knew that it would be used in this way. the fact that they said no the day before they started bombing is suspicious timing alone when they could have said something much sooner.
sakuraba@lemmy.ml · 2 pts · 155d
not sure, I think this is just anthropic trying to keep their tech closed (they already started limiting what you can do with their API)
anthropic will keep providing claude to them until they are done transitioning to stop using it in favor of other providers but it will take around 6 months to do so
quick_snail@feddit.nl · 4 pts · 154d
Palentir AI is actually anthropic. That's how anthropic helped with the kidnapping of Maduro without knowing they did
logi@lemmy.world · 2 pts · 154d
That's a big claim. Any support for that?
quick_snail@feddit.nl · 1 pts · 154d
Yeah, it's been well reported.
https://www.politico.com/news/2026/02/26/incoherent-hegseths-anthropic-ultimatum-confounds-ai-policymakers-00800135
Also they used Anthropic AI to identify targets in Iran, and many are questioning if that's why they bombed the school
https://www.wsj.com/livecoverage/iran-strikes-2026/card/u-s-strikes-in-middle-east-use-anthropic-hours-after-trump-ban-ozNO0iClZpfpL7K7ElJ2
https://thisweekinworcester.com/exclusive-ai-error-girls-school-bombing/
https://futurism.com/artificial-intelligence/pentagon-ai-claude-bombing-elementary-school