Clearly, even the “Mecha Hitler” meltdown was not sufficient to scare the Pentagon.
A number of days after Xai’s Grok Chatbot produced a stream of anti-Semitistic and homophobic reactions, Elon Musk’s firm introduced it had secured a $200 million protection contract to develop and provide AI instruments to federal staff.
The Division of Protection Chief Digital and Synthetic Intelligence Workplace introduced the contract on Monday, saying it was designed to allow the federal government to benefit from all the most recent AI instruments, from battlefield operations to enterprise wants.
“The adoption of AI is altering the division’s means to help our fighters and keep a strategic benefit over our enemies,” Dr. Doug Matty, Chief Digital and AI govt at CDAO, mentioned in an announcement. “Using off-the-shelf options in an built-in purposeful method will speed up the usage of superior AI as a part of intelligence, enterprise and enterprise data techniques as a part of the important duties of joint missions within the Warfighting area.”
In line with Lucas Hansen, co-founder of AI Safety and Training’s nonprofit Civai, the Xai deal is about rapid use and sustaining tempo with its international rivals.
“The federal government desires to help the event of frontier racebo and AI in order that the US doesn’t fall into China,” Hansen mentioned. Decryption. “Then there’s the query of how these funds are distributed and the way that help is supplied. This can be the precise approach you select to help the corporate relatively than anticipating an inside deployment quickly.”
Nonetheless, Hansen warned that Grok’s underlying actions raises deeper considerations about the way it will act in a high-stakes authorities setting.
“Even when Grok was deployed by the federal government utilizing the identical model that beforehand demonstrated his willingness to generate dangerous responses, it’s doubtless that he would behave radically unethical and autonomously, even when these use instances weren’t explicitly anti-Semitism or hatred,” he mentioned.
Grok has been on the coronary heart of a number of public meltdowns. In Might, experiences surfaced that Grok had inserted his “white genocide” claims at an unrelated immediate. Xai has condemned the issue of fraudulent workers. Final week, simply earlier than the discharge of Grok 4, the chatbot once more sparked criticism of a wave of racist and homophobic output, resulting in X CEO Linda Yaccarino’s resignation.
The priority stays about how Grok behaves in delicate environments, however Hansen mentioned the model the Pentagon makes use of shouldn’t be the identical because the model X customers have accessed since November 2023.
The $200 million award is the most recent deal between the US navy and high AI builders. Different AI builders receiving the $200 million deal embrace Google, Anthropic, and Openai. In November, Meta mentioned the open supply llama AI mannequin could be out there to US protection companies and contractors. In June, ChatGpt developer Openai was awarded a novel $200 million contract with the Division of Protection to launch the AI suite, Openai for Authorities.
If the Pentagon deployed Grok to its rival, Hansen emphasised that Grok is “openweight”, that means that it may well run on non-public {hardware} and doesn’t require an API connection again to Xai.
Grok’s Announcement for the Authorities – A set of merchandise that can allow US authorities prospects to make use of frontier fashions
We’re significantly enthusiastic about two new partnerships for US authorities companions
1) New contract from the US Division of Protection
2) We…– Xai (@xai) July 14, 2025
Hansen mentioned regardless of Grok’s latest controversy, regardless of being on the coronary heart of the commerce, they may stay near the important thing gamers shaping the way forward for AI, Hansen mentioned.
“Most of it is a authorities that wishes to have connections with all frontier rabos, and I feel it actually hit 4 vital labs,” he mentioned.
In one other assertion saying its contract with the Pentagon, Xai mentioned, “It is about creating AI instruments to help humanity looking for understanding and data. To do that, Xai introduced the launch of Grok for Authorities, a set of AI instruments, together with the just lately launched Grok 4 and Deep Search.
“America is a world chief in AI, and that is just a bit much less due to the custom of innovation and a powerful funding in engineering and science,” writes Xai. “We look ahead to with the ability to contribute to the nation that has made Xai distinctive.”
Xai didn’t reply to requests for remark Decryption.
