United Launch Alliance (ULA), a joint venture between Boeing and Lockheed Martin, is pioneering the use of OpenAI’s government-compliant artificial intelligence chatbot in the defense sector. This initiative, named “RocketGPT,” involves a trial program with approximately 150 employees.
The AI chatbot operates on Microsoft Corp.’s Azure secure cloud platform, a system approved for handling sensitive government data compliant with International Traffic in Arms Regulations (ITAR). ULA CEO Tory Bruno expressed enthusiasm, stating to SpaceNews that RocketGPT is a tool to tackle “drudgery” and “tedious, time-consuming things” in tasks such as report writing, proposal drafting, and flight telemetry analysis.
This deployment is a significant step for AI in defense, where standard consumer versions of ChatGPT are prohibited. ULA collaborated with OpenAI and PwC for months to customize RocketGPT, with around 20 people from the three organizations involved in training and security modifications. Bruno emphasized that RocketGPT serves as a research assistant, not a job replacement, stating, “AI is really, really good at handling massive volumes of data…So if that isn’t what you’re doing, that’s probably not the right tool.”
The contract represents a win for OpenAI, which recently secured a major $200 million Pentagon contract for national security AI development. While celebrating the technology, Bruno cautioned against unrealistic expectations, saying, “You don’t just grab an AI off the shelf and just turn it on…You have to train it on massive quantities of data.” He acknowledged that human oversight remains crucial due to potential AI errors. The aim, according to Bruno, is to improve worker speed and product quality, not eliminate jobs.
The pilot program includes employees across various departments, demonstrating ULA’s exploration of RocketGPT’s potential across its operations. While the investment amount remains undisclosed, Bruno described it as “relatively modest” compared to anticipated productivity increases.