Generative AI testing

NASA Tests OpenAI for Multiple Use Cases

NASA is testing OpenAI for code-writing assistance, research summarization and other applications.

NASA employees are using the Microsoft Azure cloud system to study generative artificial intelligence in a secure environment. Edward McLarney digital transformation lead for AI and machine learning at NASA, said as a precaution, employees looking to test generative AI are only allowed to apply the technology if activities involve public, non-sensitive data.

According to McLarney, employees are currently assessing OpenAI’s usability, accuracy, completeness of outputs, security behavior, speed and operating costs. NASA plans to study the provider’s chat, code assistance capabilities and image-generating capabilities, with more tests planned in the future, FedScoop reported.

The generative AI test comes months after NASA Chief Information Officer Jeff Seaton issued an agency-wide email notifying employees that while the technology is not officially cleared for use, tests will be done to identify risks and opportunities. According to Seaton, NASA is working with other federal government agencies, private industry and tech experts to understand the technology and associated policies around generative AI and how other organizations are adopting the technology.

The tests also come after the Government Accountability Office said in July that it will assess the risks posed by generative AI tools at the behest of Sens. Ed Markey and Gary Peters.