Elon Musk’s SpaceX, xAI join secret Pentagon drone swarming contest

Elon Musk’s SpaceX and its wholly owned subsidiary xAI are competing in a previously unreported Pentagon contest to develop voice‑controlled autonomous drone swarming technology, people familiar with the matter told Bloomberg. 

The entry by the two Musk companies — which he announced in February would merge — marks a new and potentially controversial move into AI‑enabled weapons development. While SpaceX is a well‑established defence contractor and Musk supports advancing AI, he has also previously argued against making “new tools for killing people.”

The companies are among only a handful selected to compete for the $100 million challenge launched in January, the sources said. SpaceX and xAI did not respond to requests for comment.

The six‑month competition aims to create advanced swarming software that can translate voice commands into digital instructions and manage multiple drones. While it’s already possible to fly several drones at once, directing them to move autonomously as a coordinated swarm — on sea and in the air — remains technically difficult. The contest is planned to progress in phases based on participants’ success.

The Pentagon initiative was launched jointly by the Defense Innovation Unit and the Defense Autonomous Warfare Group, part of U.S. Special Operations Command. The effort builds partly on the Biden‑era Replicator programme, which sought to produce large numbers of expendable autonomous drones.

A defence official said at the contest’s announcement that the drones will be used for offensive purposes, noting that the human‑machine interface “will directly impact the lethality and effectiveness of these systems.”

xAI has been hiring engineers with active U.S. security clearances to work with federal contractors, according to its website, and has signed Pentagon contracts to integrate its Grok chatbot into government sites. The company previously secured a $200 million Pentagon contract to integrate xAI into military systems.

SpaceX, a longtime defence contractor, has focused on rockets, space exploration, military communications and intelligence systems rather than offensive weapons software. Along with Boeing and Lockheed Martin it launches sensitive Pentagon satellites.

Musk has argued for banning offensive autonomous weapons that can select and engage targets without meaningful human control. In 2015 he signed an open letter from AI and robotics researchers warning against autonomous weaponry.

xAI, started in 2023, owns Musk’s AI startup, the social network X and the Grok chatbot. It recently agreed to merge with SpaceX in a deal valued at $1.25 trillion. A SpaceX statement said the acquisition will build “the most ambitious, vertically‑integrated innovation engine on (and off) Earth,” but did not mention weapons technology.

The first phase of the contest focuses on software development, progressing later to live testing. The software is intended to coordinate drone movements across multiple domains, including air and sea. Later stages will involve “target‑related awareness and sharing” and ultimately “launch to termination.”

OpenAI is also involved indirectly; it is supporting a submission from Applied Intuition. According to documents reviewed by Bloomberg, OpenAI’s contribution would be limited to converting voice and other instructions from battlefield commanders into digital instructions, and “wouldn’t be used for the operation of the drone swarm, weapons integration or targeting authority.” An OpenAI spokesperson said any use of its tools would comply with its usage policy.

By contrast, SpaceX and xAI are expected to work on the full project together.

Some defence officials have expressed alarm at integrating chatbots and voice‑to‑text AI in weapons platforms, stressing the importance of limiting generative AI to translation and keeping humans in the loop. Critics warn large language models can produce biased or unanchored outputs.

The Pentagon’s new AI Acceleration Strategy, released in January, seeks to “unleash” AI agents for battlefield use, including planning and targeting that could involve lethal strikes.

Defence contracts have previously sparked controversy in the tech industry, including protests at Google in 2018 over Project Maven, an AI initiative to analyze drone footage.

By Sabina Mammadli