The Biden administration is taking steps to ensure that developers of major artificial intelligence (AI) systems disclose the safety test results of their products to the government. The White House AI Council is scheduled to meet on Monday to review the progress made on President Joe Biden’s executive order, which was signed three months ago aiming to manage the fast-evolving technology.
One of the main mandates in the executive order is for AI companies to share vital information, including safety tests, with the Commerce Department under the Defense Production Act. This ensures that the government has access to data that ensures AI systems are safe before they are released to the public.
To achieve this, the government’s National Institute of Standards and Technology will create a uniform framework for assessing safety, as this common standard is not yet in place. However, several software companies are committed to a set of categories for the safety tests.
AI has emerged as a leading economic and national security consideration for the federal government. Given the investments and uncertainties caused by the launch of new AI tools such as ChatGPT that can generate text, images, and sounds, it is vital to regulate the safety of AI systems. The Biden administration is also looking at congressional legislation and working with other countries and the European Union on rules for managing the technology.
The Commerce Department is developing a draft rule on U.S. cloud companies that provide servers to foreign AI developers. Nine federal agencies, including the departments of Defense, Transportation, Treasury, and Health and Human Services, have completed risk assessments regarding AI’s use in critical national infrastructure, such as the electric grid.
The government seeks to ensure that the safety of AI systems is a top priority, and that companies meet this bar before releasing them to the public. Ben Buchanan, the White House special advisor on AI, added that the government wants to know companies have assessed safety effectively, as the implications of AI technology can be severe.
