OpenAI has shifted its approach by collaborating with the Pentagon on various initiatives, including the development of cybersecurity capabilities.
This marks a departure from the startup’s previous policy of refraining from providing its artificial intelligence to military entities.
Bloomberg reports that according to Anna Makanju, the company’s Vice President of Global Affairs, OpenAI is actively engaged in creating tools with the US Defense Department for open-source cybersecurity software.
The collaboration extends to participating in Defense Advanced Research Projects Agency’s (DARPA) AI Cyber Challenge, as announced the previous year, and holding preliminary discussions with the US government to explore ways to address veteran suicide prevention.
OpenAI recently revised its terms of service, eliminating language that previously prohibited the use of its AI in “military and warfare” applications. Makanju characterised this decision as part of a comprehensive policy update to accommodate new applications of ChatGPT and other tools.
Despite this change, OpenAI maintains a strict prohibition on using its technology for developing weapons, causing harm, or damaging property, according to Makanju.
Notably, Microsoft, OpenAI’s primary investor, provides software contracts to various branches of the US armed forces.
In addition to its cybersecurity efforts, OpenAI is intensifying its commitment to election security, dedicating resources to ensure that its generative AI tools do not contribute to the spread of political disinformation.
Sam Altman, OpenAI’s CEO, emphasised the significance of elections and acknowledged the prevailing anxiety surrounding them.
