I made another comment recently about how mistakes being pushed on AI systems in targeting are human decisions with human culprits. I think this is the best example of that I have seen. A civilian...
Officials involved in the investigation pointed to gaps that they said were left after Hegseth dismantled most of the Pentagon’s civilian harm mitigation, or CHM, units — cutting headcount across a number of teams by roughly 90% to fewer than 20 staff members, people with direct knowledge of the matter said. Centcom’s team was reduced from 10 to 1
I made another comment recently about how mistakes being pushed on AI systems in targeting are human decisions with human culprits. I think this is the best example of that I have seen.
A civilian harm mitigation unit is made up of intelligence officers, legal officers, and specialist support officers who are experts in kinetic effects analysis, or how a particular munition will impact a target. Together, they would assess every targeting option to figure out whether it was legal under international humanitarian law, what the impact on the target would be from a military ends perspective, collateral damage assessments, and environmental hazards, like secondary explosions from civilian gas or petroleum infrastructure.
From there, they'd present their assessment to the operations cell, and whether all the criteria imposed by law and their combatant commander were met (in this case, Commander - CENTCOM). If the criteria weren't met, they would have developed other options. These options included low-collateral damage variants of particular munitions that could mitigate risk against civilians and infrastructure.
If they could not get the risk level down to the combatant commander's standing orders, the strike would be canceled.
Other operational military elements would be considered, like the insertion of special forces, if the operational plan required human judgement, and if it was vital that the particular target or target be damaged, degraded, or destroyed.
'That's a fuckin school' would have taken one intelligence analyst five minutes, accessing all-source data that was already on file. It is not a hard determination to make. The legal officer then would have marked the targeting package illegal due to breaching the principle of distinction.
An operations center during a large scale air-focused operation is extremely busy. 20 staff members across the entire Pentagon for a war against an enormous country is a deliberate choice to diregard the safeguards protecting civilian deaths.
These are not AI failures. These are deliberately cruel and illegal acts, carried out by human beings. The US has always been against the International Criminal Court due to looser interpretations of international humanitarian law compared to many of its NATO allies.
This was such a known risk during the 2002 Iraq War that the US passed the Hague Invasion Act to violate the sovereignty of the Netherlands and 'protect' US service members who had been arrested for war crimes.
The recent attacks on international institutions, and specifically the International Criminal Court, make much more sense viewed through this lens. The humans who made these decisions to not discriminate against military and civilian targets (and to remove risk mitigations against knowingly breaking international law) would be very easy targets for prosecution, were they not American.
So if I'm understanding, and this sounds crazy, but when you dismantle the "don't do war crimes" unit, no one stops you from doing the war crimes*? It almost** sounds deliberate! Sorry for the...
So if I'm understanding, and this sounds crazy, but when you dismantle the "don't do war crimes" unit, no one stops you from doing the war crimes*? It almost** sounds deliberate!
Sorry for the sarcasm but I just think if I don't mock it, I'll cry.***
*War crimes may vary.
** Almost here means "absolutely"
***Tears guaranteed anyway
From the article: [...] [...] [...] [...] [...] [...]
From the article:
Within hours of a Feb. 28 missile attack that hit an Iranian school, some Pentagon personnel knew the US was responsible, according to officials involved in an internal investigation of the incident. That probe found what the officials now describe as a cascade of preventable failures.
[...]
United Nations investigators said Thursday that there were reasonable grounds to conclude that the strike on Minab and another US attack that took place on the same day amounted to war crimes. The US has yet to publicly accept responsibility for the school strike, and President Donald Trump has previously suggested it was Tehran’s fault. As recently as July 14, he told Fox News, “I don’t think anybody’s going to ever be able to say what happened there.”
But in interviews, officials with direct involvement in the Pentagon’s internal investigation have offered some of the first accounts of the probe. The strike didn’t result from a single catastrophic decision but an accumulation of smaller ones, they said, and their review has identified a series of missed opportunities to prevent the disaster. The officials, like more than two dozen others who provided descriptions of targeting procedures or background information for this story, asked not to be named to discuss sensitive matters.
[...]
A report on the full Pentagon investigation, which commenced in March, hasn’t been released — though the officials said it has been all-but-complete for several months.
[...]
US officials have said that outdated intelligence was a problem. Although the site in Minab had been identified years earlier as a military compound, there was clear evidence that it had been modified and undergone a change of purpose, the officials said. Physical changes reflecting new construction — visible on satellite imagery — dated back almost a decade, they said, and the site operated openly as a school.
[...]
Some US intelligence analysts took notice, the officials said. Bloomberg reported in June that one analyst detected changes to the site as early as 2019, and logged remarks about the modifications in a system that wasn’t connected to the primary military intelligence database, which informs targeting.
Inside Centcom, which conducted the US attack, some personnel relied too much on the artificial intelligence embedded in Maven Smart System, the officials said. Maven, which is made by Palantir Technologies Inc., allows users to view and coordinate complex military operations — from targeting to command and control — using AI-enabled software that fuses more than 150 different data inputs into a coherent set of information to inform commanders’ decisions.
In the past year, the Defense Department has rapidly made Maven a cornerstone of the US military apparatus. Several former senior US military officials now work for Palantir, including some with high-level security clearances in Centcom’s Tampa headquarters. Their role is to keep the software functioning inside secure military environments.
Some Centcom personnel expected Maven to flag stale information or inconsistencies in the underlying intelligence assembled for potential targets, the officials said. It’s not clear why they had such expectations.
[...]
After the Minab strike, Palantir built new capabilities into Maven that “re-review underlying intelligence to identify factors that would disqualify a target and flag inconsistencies and inaccuracies that human review may have missed,” according to a person familiar with the matter. That work has already caught some anomalies, the person said.
[...]
In past conflicts, preparing lists of proposed targets — work that includes assessing how they align with an operation’s objectives — had taken hours for staffers to complete. Ahead of the US attack on Iran, much of that work was condensed into several minutes via Maven, people involved in the Pentagon’s internal review said.
Officials involved in the investigation pointed to gaps that they said were left after Hegseth dismantled most of the Pentagon’s civilian harm mitigation, or CHM, units — cutting headcount across a number of teams by roughly 90% to fewer than 20 staff members, people with direct knowledge of the matter said. Centcom’s team was reduced from 10 to 1.
No CHM team member reviewed the Minab site before the strike, according to officials involved in the internal investigation. That was a consequence of the cuts in staffing but also a decision not to involve the group in planning and operations, the officials said. While not mandatory, that sort of review had become routine in recent years, they added. It’s unclear whether such checks, which include mapping the civilian environment, assessing noncombatant presence and developing lower-risk strike options, would have been enough to raise questions about the Minab site.
I made another comment recently about how mistakes being pushed on AI systems in targeting are human decisions with human culprits. I think this is the best example of that I have seen.
A civilian harm mitigation unit is made up of intelligence officers, legal officers, and specialist support officers who are experts in kinetic effects analysis, or how a particular munition will impact a target. Together, they would assess every targeting option to figure out whether it was legal under international humanitarian law, what the impact on the target would be from a military ends perspective, collateral damage assessments, and environmental hazards, like secondary explosions from civilian gas or petroleum infrastructure.
From there, they'd present their assessment to the operations cell, and whether all the criteria imposed by law and their combatant commander were met (in this case, Commander - CENTCOM). If the criteria weren't met, they would have developed other options. These options included low-collateral damage variants of particular munitions that could mitigate risk against civilians and infrastructure.
If they could not get the risk level down to the combatant commander's standing orders, the strike would be canceled.
Other operational military elements would be considered, like the insertion of special forces, if the operational plan required human judgement, and if it was vital that the particular target or target be damaged, degraded, or destroyed.
'That's a fuckin school' would have taken one intelligence analyst five minutes, accessing all-source data that was already on file. It is not a hard determination to make. The legal officer then would have marked the targeting package illegal due to breaching the principle of distinction.
An operations center during a large scale air-focused operation is extremely busy. 20 staff members across the entire Pentagon for a war against an enormous country is a deliberate choice to diregard the safeguards protecting civilian deaths.
These are not AI failures. These are deliberately cruel and illegal acts, carried out by human beings. The US has always been against the International Criminal Court due to looser interpretations of international humanitarian law compared to many of its NATO allies.
This was such a known risk during the 2002 Iraq War that the US passed the Hague Invasion Act to violate the sovereignty of the Netherlands and 'protect' US service members who had been arrested for war crimes.
The recent attacks on international institutions, and specifically the International Criminal Court, make much more sense viewed through this lens. The humans who made these decisions to not discriminate against military and civilian targets (and to remove risk mitigations against knowingly breaking international law) would be very easy targets for prosecution, were they not American.
So if I'm understanding, and this sounds crazy, but when you dismantle the "don't do war crimes" unit, no one stops you from doing the war crimes*? It almost** sounds deliberate!
Sorry for the sarcasm but I just think if I don't mock it, I'll cry.***
*War crimes may vary. ** Almost here means "absolutely" ***Tears guaranteed anywayFrom the article:
[...]
[...]
[...]
[...]
[...]
[...]