Menu Menu
[gtranslate]

US government strikes $200m Grok deal to modernise Defense Department

A week after Grok’s ‘MechaHitler’ debacle, the US government has announced a $200m contract with the AI platform to modernise the Defense Department.

The chaotic relationship between Donald Trump and Elon Musk is throwing out some ridiculous headlines, and this is just the latest.

If you’re chronically online, like me, you’ll be familiar with Grok’s peculiar crash out last week, in which the chatbot anointed itself ‘MechaHitler’ and generated a slew of antisimetic posts on X. We’re aware Musk is a stalwart of free speech, but damn.

As quickly (and randomly) as it began, the chatbot’s dark comedy era appears to now be over; the posts have been removed, and company engineers are presumably peering into the bot’s code to find out what the hell prompted the spiral?

Amid all the lingering confusion and public distaste, what better time is there to announce a major deal between the potty mouth chatbot and the US government? It’ been a whole week, to be fair.

Musk’s xAI has just been granted $200m from the Pentagon to modernise the nation’s Defense Department. This follows several similar contracts penned with Google, Anthropic, and OpenAI in recent months.

While details from the government’s announcement are sparse – given the clandestine nature of national security – xAI has talked up a package it calls ‘Grok For Government’ which will provide AI tools to ‘federal, local, state, and national security customers’.

‘These customers will be able to use the Grok family of products to accelerate America – from making everyday government services faster and more efficient to using AI to address unsolved problems in fundamental science and technology,’ a statement read.

It’s a little (or a lot) concerning to think generative AI could soon underpin military strategies and matters of national security, given the tycoons responsible for their inception are taken aback by their proclivity to go rogue unexpectedly.

Grok’s haywire half-day last week was apparently traced back to an update where it had been instructed not to be ‘afraid to offend people who are politically correct.’ However, it had not been anticipated that Grok would ‘ignore its core values’ in order to generate ‘engaging responses for users’ across that 16 hour period.

Between celebrating the deaths of ‘future fascists in Texas’ and dubbing itself ‘MechaHitler’, users were certainly engaged.

AI is still developing at a rate faster than we’re able to effectively regulate, and the notion of feeding large language models private data is arguably too high a stake for any potential benefit they may offer right now. The thought of government officials using generative AI to supplement critical thinking is even scarier, especially when considering who owns these private companies.

Musk is hardly impartial when it comes to approaching US legislation. Just look at his recent public unravelling following Trump’s announcement of the One Big Beautiful Bill, and its cull on electric vehicle subsidies. If public approval polls are anything to go by, Musk is a polarising figure at best, and many will doubt the trustworthiness of any tool he has a hand in creating.

Ethics watchdogs, lawmakers, and private advocates are growing increasingly uncomfortable with billionaires monopolising public infrastructure, and Musk already has his tendrils wrapped around space travel, transportation, communication technologies, and AI. Does he really need to ‘upskill’ government, armed forces, and healthcare?

For there to be any sense of civic calm around the situation, third party regulation is essential to ensure democratic accountability. It doesn’t appear that’s part of the plan though.

Accessibility