OpenAI is partnering with defense startup Anduril Industries to develop AI for the Pentagon. On Wednesday, the companies announced that they will combine OpenAI models, including GPT-4o and OpenAI o1, with Anduril systems and software to improve the US military’s defense against unmanned aerial attacks.
The deal comes less than a year after OpenAI softened its stance on the use of its models for military purposes. While the ChatGPT maker’s policy still prohibits the use of its models for weapons development or use, in January it removed a line that explicitly banned the integration of its technology into “military and warfare activities.” At the time, the company said it was already working with DARPA on cybersecurity tools. In October, the company hired a former Palantir security officer and was reportedly going to present its products to the US military and national security agencies.
A spokesperson for OpenAI told The Washington Post that the deal is in line with the company’s rules, as it focuses on systems that protect against threats from unmanned aerial vehicles. The company said that the partnership does not apply to other areas of use.
According to The Washington Post, the partnership between OpenAI and Anduril will be aimed at improving technologies for detecting and shooting down drones that threaten the US military and its allies. The Pentagon is already purchasing a Roadrunner drone interceptor (pictured above) from Anduril to help counter the growing number of smaller drones on the world’s battlefields. The startup sells watchtowers, communication jammers, military drones, and an autonomous submarine, among other projects.
The companies have framed the partnership as a way to protect U.S. military personnel and counter the development of Chinese artificial intelligence. “Our partnership with OpenAI will allow us to leverage their global expertise in artificial intelligence to address pressing air defense challenges around the world,” Anduril CEO Brian Schimpf wrote in a statement. “Together, we are committed to developing responsible solutions that will enable military and intelligence services to make faster, more accurate decisions in high-pressure situations.”
Anduril was co-founded by Oculus Rift inventor (and co-founder of Oculus VR) Palmer Luckey. This headset laid the foundation for the Meta Quest line, which today holds the lion’s share of the VR and AR market. Lucky left Meta (then Facebook) in 2017, a few months after news broke that he had donated $10,000 to a group that aimed to post 4chan-style memes against Hillary Clinton on roadside billboards.
“OpenAI builds AI to benefit as many people as possible, and supports U.S. efforts to ensure that technology supports democratic values,” OpenAI CEO Sam Altman wrote in a statement. “Our partnership with Anduril will help ensure that OpenAI’s technology protects U.S. military personnel, and help the national security community understand and use this technology responsibly to ensure the safety and freedom of our citizens.”
