Pentagon AI chief: Musk’s Grok chatbot used to launch thousands of missiles at Iran

The Hill ^ | June 17th, 2026 | by Ashleigh Fields

[Mariner]

The Pentagon artificial intelligence chief on Monday said Elon Musk’s Grok chatbot is tantamount to national security in a sworn statement that noted xAI’s technology has been used throughout the Iran war.

Cameron Stanley, the chief digital and artificial intelligence officer for the Department of Defense, wrote that the chatbot “enabled U.S. forces to deploy over 2,000 munitions to 2,000 distinct targets within 96 hours during Operation Epic Fury, a testament to the greatly increased operational efficiency made possible by the Grok Gov Model.”

Stanley’s statements were used as evidence in efforts by the Trump administration to preserve the xAi data center near Memphis, where the NAACP alleges the company is illegally polluting the air.

The Justice Department has asked the judge to throw out the case, citing the need for access to the Colossus 2 data center, which Stanley described as “vital” to national security missions, including targeting, intelligence, readiness and recruitment.

“Furthermore, it has tailored functionality to support military planning workflows, report synthesis and generation, predictive analytics for logistics and sustainment, red-teaming analysis of adversary positioning, personnel management, and medical supply lines,” Stanley wrote.

---

TOPICS: Foreign Affairs; News/Current Events
KEYWORDS:

[Mariner]

2,000 sorties against 2,000 distinct targets in 96 hours.

Is that extraordinary?

Is there a Mission Planner in the crowd?

[CatOwner]

Why would this need to be public knowledge?

[Mariner]

Pentagon AI Chief thought it should be.

Or his boss did.

[Brian Griffin]

“An initial survey released by two nonprofits earlier this week of air pollution collected from community-run air monitors at three sites throughout southwest Memphis shows that PM2.5 levels were consistently above EPA limits between November 2025 and March of this year.”

https://www.motherjones.com/politics/2026/06/xai-data-center-neighbors-angry-spacex-ipo-gas-turbines-colossus-pm2-5-elon-musk-trillionaire-pollution-asthma/

[Mariner]

AI Overview

Colossus 2, Elon Musk’s xAI supercomputer in Memphis/Southaven, operates at a massive 1 to 1.5 gigawatt (GW) power scale. This colossal energy footprint—equivalent to powering a major city like San Diego—is supplied by an on-site private power plant relying heavily on a fleet of mobile natural gas turbines and Tesla Megapacks for stability.

[Gunslingr3]

2,000 sorties against 2,000 distinct targets in 96 hours. Is that extraordinary?

In the first 24 hours (Day 1, starting ~02:38 Baghdad time on 17 January 1991), the coalition flew approximately 2,775 sorties (including strikes and support), attacking a large number of strategic targets (over 150 discrete ones in the initial master attack plan).

For the first 96 hours (roughly the first four days of the air campaign), reliable aggregate figures are harder to pinpoint in unclassified public sources, but contemporary analyses and summaries indicate around 7,000+ combat ("shooter") sorties were flown in that period, with heavy emphasis on suppressing air defenses, command/control, and other strategic targets.

The opening phase prioritized Iraqi air defenses (KARI system), airfields, command centers, Scuds, and other high-value targets. F-117s, cruise missiles (Tomahawks and CALCMs), and other assets played key roles initially.

Sortie rates were very high early on (averaging well over 1,000 per day initially) before settling into a more sustained pace. Total Desert Storm air sorties exceeded 100,000–116,000 over ~5–6 weeks (with ~40,000 strike and ~50,000 support), so the first four days represented an intense opening surge.

[ckilmer]

yes. the vast number of precisely targeted sorties set up in real time was the big upgrade that ai enabled.

likely the raid that was done before the peace deal was an even more upgraded version of that.

If the US and israel have to go in again next year the ai intelligence on Iran will be even more powerful

[Bratch]

What could possibly go wrong?

[captmar-vell]

“Why would this need to be public knowledge?”

Its a dig against Anthropic, which tried to limit what our Military was able to do and was summarily fired. Then they had an assortment of curious mishaps with their source code being ‘accidently’ released, certainly the CCP and Russia both have it now (I’m guessing)

They are also about to IPO at $65 Billion.

Our evil woke Tech lords think they know best. I’m not buying what they’re selling, or their IPO.

[Bob Wills is still the king]

“Siri - Please launch missile barrage at Tehran...”