The Pentagon has launched a $100m (£74m) competition to develop an AI bot that can be used to translate voice or written commands from soldiers to a fleet of drones.

Mr Musk’s SpaceX is one of the companies pitching for a share of the work, Bloomberg reported.

OpenAI, the developer of ChatGPT, is also said to be working with US autonomous vehicles business, Applied Intuition, on a rival bid.

The US military’s Defence Innovation Unit last month confirmed it was seeking bidders to develop an “autonomous vehicle orchestrator”, an AI system that can be used by ordinary soldiers in the field to command “autonomous systems at the fleet level” that can be used to “overwhelm our adversaries”.

Portable drones have proven to be vital weapons on the frontlines in Ukraine and are used for long-range surveillance and lethal attacks while operated by pilots holed up miles away.

Portable drones have proven to be vital weapons for Ukraine during its war with Russia

Drone swarms are seen as crucial hardware for soldiers in future conflicts. They could be used to swamp enemy defences or guard friendly troops.

Lt Gen Frank Donovan, the director of the US military’s Defence Autonomous Warfare Group, said the system would “allow humans to work the way they already command – through plain language … not by clicking through menus or programming behaviours”.

He added: “We must ensure the human – the ethical decision maker – always maintains a clear understanding of what the system is doing and why.”

Mr Musk’s SpaceX is already a key defence contractor for the US government, providing rocket launches and communications through its Starlink satellite business.

Efforts to bid for the military contract come after Mr Musk merged his SpaceX business with xAI, another company he owns and which controls social media platform X and the AI bot Grok.

‘Pandora’s box’

Mr Musk has previously warned against the prospect of AI-powered weapons. In 2017, he signed a letter warning that AI warfare threatened to be a “Pandora’s box” and in 2018 cautioned that AI was “far more dangerous than nukes”.

Silicon Valley AI labs have jumped at the opportunity to scoop up lucrative US defence contracts and offer up their tools for military use. In 2024, OpenAI, the developer of ChatGPT, scrubbed a ban on military use of its AI technology from its company policies.

Earlier this month, OpenAI said it would be offering a customised version of ChatGPT to the Pentagon. A spokesman said: “With the proper safeguards, AI can help protect people, deter adversaries, and prevent future conflict.”

Story Continues

OpenAI has a $200m contract to work with the US department of defence’s AI office. Mr Musk’s xAI, Google and AI lab Anthropic all have similar deals to provide their AI bots to the US military.

However Anthropic, the developer of the Claude chatbot, is reported to have clashed with the Pentagon over the military’s plans for its technology.

The US has been urging companies to allow their AI to be deployed for “all lawful purposes”, something Anthropic has been resisting, Axios first reported.

SpaceX, OpenAI and Applied Intuition were contacted for comment.

Terms and Privacy Policy

Privacy dashboard

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.