Briefly
- The Pentagon plans to deploy AI techniques throughout labeled and unclassified navy networks.
- xAI’s Grok will likely be added to the Pentagon’s GenAI.mil platform alongside Google’s Gemini.
- Public Citizen warned that Grok’s security file poses nationwide safety dangers.
Secretary of Protection Pete Hegseth stated Monday that the Pentagon is shifting to make synthetic intelligence a core a part of the U.S. navy’s functionality, utilizing a speech at SpaceX’s Starbase facility in Texas to stipulate how AI and house applied sciences will form future operations.
Hegseth stated the Pentagon’s goal is to turn out to be “an AI-first warfighting power throughout all domains,” together with each inner planning techniques and frontline operations.
“Merely put, the US should win the strategic competitors for twenty first century technological supremacy,” Hegseth stated, highlighting synthetic intelligence, autonomous techniques, quantum, hypersonics, and long-range drones. “When you speak to Elon Musk lengthy sufficient, he’ll inform you how vital hypersonics and long-range drones are, and he is 100% appropriate. Area capabilities, directed power, and biotechnology are the brand new areas of worldwide competitors.”
Hegseth stated AI will transfer shortly into broad operational use throughout the Protection Division.
“Very quickly, we may have the world’s main AI fashions on each unclassified and labeled community all through our division,” he stated, calling it “lengthy overdue.”
In a nod to the renewed relationship between the Trump Administration and Musk, Hegseth stated the Pentagon would start utilizing Grok.
“As we speak, we’re excited to announce the following frontier AI mannequin firm to affix GenAI.mil, and that’s Grok from xAI, which is able to go reside later this month,” Hegseth stated.
The xAI deal locations Grok inside a rising ecosystem of business AI instruments already used throughout the federal authorities, alongside fashions from OpenAI, Google, Anthropic, and Microsoft. By way of Pentagon contracts and GSA-approved agreements, these AIs are getting used for protection preparation, battle simulation, intelligence gathering, and information evaluation. Hegseth stated the intention is to take care of U.S. navy benefit as AI capabilities unfold globally.
“We should guarantee America’s navy AI dominance in order that no adversary can exploit that very same expertise to carry our nationwide safety pursuits or our residents in danger,” he stated.
He criticized present protection acquisition timelines as misaligned with fashionable battle.
“In fashionable warfare, the quickest innovator and iterator would be the winner,” Hegseth stated. “We’re performed operating a peacetime science truthful whereas our potential adversaries are operating a wartime arms race.”
Forward of Hegseth’s remarks, SpaceX CEO Elon Musk described Starbase as a part of a long-term effort to push human spaceflight past science fiction.
“We need to make Star Trek actual,” Musk stated. “We need to make Starfleet Academy actual in order that it’s not all the time science fiction, however at some point the science fiction turns to science truth.” He stated SpaceX’s aim is to construct “massive spaceships with folks going to different planets, going to the moon, and finally going past our star system.”
Whereas Musk targeted on how SpaceX would convey the dream of deep house exploration to life, the choice to combine Grok into Pentagon techniques has drawn criticism from civil society teams.
In an announcement, J.B. Department, Large Tech Accountability Advocate at Public Citizen, stated Grok’s deployment contained in the Protection Division poses severe dangers.
“Permitting an AI system with Grok’s observe file of repeatedly producing nonconsensual sexualized pictures of ladies and kids to entry labeled navy or delicate authorities information raises profound nationwide safety, civil rights, and public security issues,” Department stated.
“Deploying Grok throughout different areas of the federal authorities is worrying sufficient, however selecting to make use of it on the Pentagon is a nationwide safety shame,” the assertion continued.
In November, Public Citizen joined greater than 30 civil rights, client safety, and expertise accountability organizations in letters urging federal businesses to halt Grok’s use throughout authorities, citing security failures, lack of transparency, and an incapacity to satisfy minimal requirements for delicate deployments.
Hegseth didn’t handle these criticisms; as a substitute, he continued to stipulate how the Pentagon will assess AI techniques.
“Accountable AI on the Battle Division means objectively truthful AI capabilities employed securely and inside the legal guidelines governing the actions of the division,” he stated. “We are going to decide AI fashions on this commonplace alone.”
The Pentagon additionally known as for restructuring of its information infrastructure, remodeling its Advana platform right into a Battle Information Platform meant to increase entry to operational information for AI techniques supporting multi-domain operations, together with house.
“We’ll closely leverage President Trump’s tech power initiative to usher in the very best and brightest from business and academia,” Hegseth stated. “With folks like Elon, David Sacks, and others from the entrepreneurial and enterprise world already in authorities, we’ve got proven that we will and that we should enlist the world’s main expertise on this trigger.”
Usually Clever Publication
A weekly AI journey narrated by Gen, a generative AI mannequin.