Tech companies are largely applauding the new regulation, which seeks to govern how the federal government will use AI and establish guidelines for companies building new models.
President Biden signed a sweeping new executive order on Monday to place guardrails on the use and development of AI, including provisions that will make large upcoming AI models like OpenAI’s GPT-5 subject to oversight before they are released.
Speaking to a room of lawmakers, industry leaders and reporters at the White House on Monday, Biden described how the executive order was designed to mitigate the risks from AI while still tapping into its benefits. “I’m determined to do everything in my power to promote and demand responsible innovation,” Biden said, calling AI the “most consequential technology of our time.”
As part of the executive order, any company building an AI model that could pose a risk to national security must disclose it to the government and share data about what is being done to secure it in accordance with federal standards to be developed by the National Institute of Standards and Technology. The decree to share pre-release testing data applies only to models that haven’t been released yet — which would include GPT-5, the much anticipated successor to hugely popular GPT-4.
“Companies must tell the government about the large scale AI systems they’re developing and share rigorous independent test results to prove they pose no national security or safety risk to the American people,” Biden said at the event.
Ben Buchanan, AI special advisor to the White House, told Forbes that models that are currently in use, like GPT-4 or Google’s Bard, are still subject to the other elements of the executive order, including “equity provisions, discrimination, protecting consumers, workers,” he said. So far, though, he added, “we’ve not seen a Chat GPT-4 enabled catastrophe that I know of.”
The order also aims to kick off a hiring blitz for AI workers in the federal government with “dozens to hundreds” of AI-focused hires, Buchanan said. Plus, it says it will reduce the barriers to immigration for international workers in the AI sector. That doesn’t include increasing the cap on the number of H1B visas, Buchanan said, but he noted there will be greater emphasis on making the overall visa process smoother for people working on “critical emerging technologies.”
The order also establishes the creation of guidelines and standards for the use of AI by the government. Addressing fears that AI could be used to discriminate against citizens, target critical infrastructure or be used in warfare, the executive order will also require large AI models and programs to be assessed by federal agencies before being deployed. Federal agencies — from the Department of Defense to the Justice Department — will also need to produce studies that outline how they plan to incorporate AI into their functions. Some provisions relating to security issues are expected to come into effect within the next 90 days.