OpenAI Now Wants to Provide AI Solutions For Military Problems

OpenAI reportedly has taken a substantial step into the military domain this week.

In a joint statement released on Wednesday, OpenAI and defense technology startup Anduril Industries announced that they will establish a strategic partnership to develop and responsibly deploy advanced artificial intelligence (AI) solutions for national security missions.

This marks OpenAI's first collaboration with a commercial weapons manufacturer and is also the deepest cooperation the company has had with the U.S. Department of Defense to date. Anduril announced a $200 million contract in November last year to provide the U.S. Marine Corps with the company's anti-drone system.

The two companies stated that the collaboration will focus on enhancing the United States' Counter-Unmanned Aerial Systems (CUAS) and its capabilities for real-time detection, assessment, and response to potential lethal aerial threats.

Anduril and OpenAI will explore how to utilize cutting-edge AI models to rapidly synthesize time-sensitive data, alleviate the burden on human operators, and improve situational awareness. These models will be trained in Anduril's industry-leading CUAS threat and action database to help protect U.S. and allied military personnel and ensure mission success.

Founded in 2017, Anduril is dedicated to developing and deploying integrated autonomous solutions for various sensors and has experience in deploying automated robotic systems in tactical environments. The company is currently valued at around $14 billion and is one of the few successes among many emerging defense startups.

Anduril Industries co-founder and CEO Brian Schimpf said, The defense solutions built by Anduril meet the urgent combat needs of the United States and its allies. The collaboration with OpenAI will enable us to leverage their world-class expertise in artificial intelligence to address the global gap in air defense capabilities. We are committed to developing responsible solutions that allow military and intelligence personnel to make faster and more accurate decisions under high-pressure conditions.

OpenAI CEO Sam Altman pointed out in the statement that OpenAI builds artificial intelligence to benefit as many people as possible and ensure that the technology upholds democratic values. Our collaboration with Anduril will help ensure that OpenAI technology protects U.S. military personnel and will assist the national security sector in understanding and using this technology responsibly to safeguard the safety and freedom of our citizens.

It is worth mentioning that Silicon Valley was once reluctant to engage with the Pentagon, but now, with the development of AI technology, the two seem to be establishing a deeper connection.

As previously reported, Meta also indicated in early November that it would allow U.S. government agencies and contractors working on national security to use the company's AI model Llama for military purposes. This move undoubtedly changed Meta's previous policy of prohibiting the use of its AI technology for military purposes.

Meta's spokesperson also mentioned at the time that the company would share its technology with other members of the Five Eyes alliance. In addition to the United States, the Five Eyes countries include Canada, the United Kingdom, Australia, and New Zealand.

OpenAI's latest shift may not have been sudden either.

In fact, industry insiders noticed as early as the beginning of this year that OpenAI had changed its policy on cooperation with the military. Initially, OpenAI prohibit the use of its large language models for any military or war-related applications. However, by January of this year, the company removed the regulation of prohibition in military and warfare from its usage policy, stating only that OpenAI's products cannot be used to harm oneself or others, including harm caused by the development of weapons, and allowed some cooperation with the military.

Although OpenAI still prohibits the use of its technology for offensive weapons, it has reached some agreements with the U.S. Department of Defense on cybersecurity work and other projects. OpenAI also hired former U.S. National Security Agency Director Paul Nakasone to join the company's board of directors and hired former Department of Defense official Sasha Baker to form a team focused on national security policy.

Other Silicon Valley tech companies are taking similar steps. AI startup Anthropic said last month that it would open up the company's AI technology to the U.S. military through a partnership with Palantir Technologies.

Industry insiders have stated that with the rise of electronic warfare, AI may become increasingly important for ensuring that drones can perform tasks in the air. Electronic warfare uses jammers to block the GPS signals and radio frequencies that drones rely on for flight. AI can also help soldiers and military commanders analyze large amounts of battlefield data.

Although neither OpenAI nor Anduril disclosed any financial information in their statement, it is expected that delving deeper into the defense sector could also open up another source of revenue for OpenAI—the company is currently seeking to transition from its initial non-profit laboratory to becoming a profitable leader in the AI industry.