OpenAI is working with the Pentagon on several projects, including those related to cybersecurity, breaking the startup's previous ban on providing AI technology to the military.
ChatGPT, a maker of artificial intelligence chatbots, is working with the US Department of Defense to develop tools around open source cybersecurity software, Anna Makango, the company's vice president of global affairs, told Bloomberg.
The company has begun preliminary discussions with the US government about ways to prevent suicide among veterans.
OpenAI previously removed a clause from its terms of service prohibiting the use of its AI models in military and war zones, raising concerns among AI safety advocates.
Makango described the move as part of a broader update to the company's policies to reflect new uses of ChatGPT and its other tools.
Makango said OpenAI insists its technology will not be used to develop weapons, destroy property or harm people.
“There are national security use cases that align with our mission,” a company spokesperson said. “We are working with DARPA to facilitate the development of new cybersecurity tools to protect critical infrastructure and open source software that industry depends on.”
“It was not clear whether our previous policy allowed for these beneficial use cases. “So the goal of updating our policy is to provide clarity and opportunity to have these discussions,” he added.
Microsoft is a major investor in OpenAI and awards numerous software contracts to the US Armed Forces and other government agencies.
Cybersecurity breaches are a major problem in many industries, as companies use artificial intelligence to combat threats.
OpenAI also announced that it is accelerating its work on election security and allocating resources to ensure its generative AI tools are not used to spread political misinformation.
The company proactively anticipates and prevents potential violations, increasing transparency of image provenance and facilitating access to reliable voting information.