OpenAI wins $200 mn contract with Pentagon to pilot AI for defence tasks

By Rachel Metz

 

OpenAI has won a $200 million contract for a pilot program aimed at helping the US Defense Department determine how it could use artificial intelligence for a range of administrative and security tasks. 

The one-year contract, which the Defense Department disclosed Monday, is the latest push by the ChatGPT-maker to sell its technology to the US government. 

 

In a blog post Monday, OpenAI said the contract is the company’s first project under a new entity it’s calling OpenAI for Government.  

 

As part of the effort, OpenAI will work with the Defense Department to come up with ways that AI can help with administrative tasks, such as getting health care for US military members and helping prevent cyberattacks.

 

OpenAI for Government consolidates all of OpenAI’s existing government projects in one area of the company, including ChatGPT Gov — a version of ChatGPT meant for government workers — as well as its work with Nasa, the National Institutes of Health, the Air Force Research Laboratory and the Treasury Department.

 

The company also said late last year that it would partner with weapons maker Anduril Industries Inc. to build AI for anti-drone systems, in another sign of its expanding work with the US government, particularly around national security. 

 

In recent months, OpenAI has said that it plans to build out AI tools that uphold democratic values. The AI startup has also brought in a former top Pentagon official to lead its national security policy team and added the former head of the National Security Agency to its board.