President Biden issued a highly anticipated, sweeping executive order on artificial intelligence (AI) on Monday, focused on seizing on the emerging technology and managing its risks.
The order includes several new actions, which focus on areas like safety, privacy, protecting workers, and protecting innovation.
“President Biden believes that we have an obligation to harness the power of AI for good, while protecting people from its potentially profound risks,” a senior administration official told reporters.
The executive order includes new standards for safety, including requiring companies developing models that pose a serious risk to national security, economic security or public health to notify the federal government when training the model, and they must share the results of all safety tests. The Commerce Department will also develop guidance for content authentication and watermarking to label AI-generated content.
The order directs federal agencies to accelerate the development of techniques so AI systems can be trained while preserving the privacy of the training data. The order will also evaluate how agencies collect and use commercially available information containing personal data.
To support workers, the order develops principles and best practices to mitigate the harms and maximize the benefits AI creates for workers by addressing issues including job displacement, labor standards, and data collection.
The order also aims to promote innovation and competition through a pilot of a tool, the National AI Research Resource, that will expand grants for AI research in areas such as health care and climate change. It will also use existing authorities to expand the ability of highly skilled immigrants and nonimmigrants with expertise in these critical areas to study, stay, and work in the U.S. by streamlining visa criteria and interviews.
The order applies to companies with the most powerful AI systems, regardless of whether the companies work with the federal government, according to an official. It marks the most significant effort to impose national order on the AI industry after ChatGPT, which launched less than a year ago, shocked users with its human-like capabilities and knowledge.
The executive order aims to have the U.S. lead the way on AI and advance American leadership globally on the emerging technology. Vice President Harris is traveling to the United Kingdom on Tuesday to give a speech on the U.S. vision for AI.
Officials argue though that the U.S. is not behind Europe in regulating AI.
“I don’t think we’re in a race, I don’t think we’re playing catch up,” the official said. “I would pushback on any notion that we are behind anyone. But … at the end, I think this is an opportunity to work together on AI governance. I don’t particularly see an amount of tension between … the United States and Europe on this.”
The order focuses on advancing equity and civil rights by providing guidance to landlords, federal benefits programs and federal contractors to keep AI algorithms from exacerbating discrimination. And it aims to develop a set of best practices for AI use on the criminal justice system, including sentencing; parole and probation; surveillance; crime forecasting; and forensic analysis.
The order also directs the Department of Health and Human Services to establish a safety program so it can receive reports of unsafe healthcare practices involving AI, and it will create resources to support educators deploying AI tools for learning.
“President Biden is rolling out the strongest set of actions any government in the world has ever taken on AI safety, security, and trust. It’s the next step in an aggressive strategy to do everything on all fronts to harness the benefits of AI and mitigate the risks,” deputy chief of staff Bruce Reed said in a statement.
The White House will appoint an AI council for principal level meetings, spearheaded by Reed, to deliver on the goals of the order, the official said. The timing of implementing the executive order, which comes nearly a year out from Election Day, is within 90 days for some of the safety actions, and 270 days or a year for other deliverables.
Additionally, the White House is launching AI.gov, a website for job applicants who want to work on AI issues in the federal government. The official said they want to bring AI experts into the federal workforce because “you can never have too much AI talent.”