Biden Administration Introduce First US AI Safety Institute Director
U.S. Secretary of Commerce Gina Raimondo announced this week that White House economic policy advisor Elizabeth Kelly will serve as the first director of the new AI Safety Institute (AISI) established by President Biden in the executive order signed in October to create government oversight of AI development and deployment. Kelly was part of the core team that drafted the order, which places AISI within the National Institute for Standards and Technology (NIST) at the Department of Commerce. NIST chief AI advisor Dr. Elham Tabassi will join Kelly as the AISI’s inaugural CTO after leading the development of the widely-used NIST AI Risk Management Framework.
In her role as AISI director, Kelly will lead and oversee operations at the institute while coordinating with other governmental AI policy and technical programs. Her term will likely reflect how she spearheaded the domestic aspect of the AI executive order, whose stated aim is encouraging competition while protecting workers and individual privacy. Meanwhile, Tabassi’s CTO position will center on core technical work around AI safety, security, and trustworthiness. Her ideas will set the tone for research guidance and model evaluations to identify and mitigate any risks from AI, especially generative AI and large language models (LLMs). That’s fitting after the success of the NIST AI Risk Management Framework.
“The Safety Institute’s ambitious mandate to develop guidelines, evaluate models, and pursue fundamental research will be vital to addressing the risks and seizing the opportunities of AI. I am thrilled to work with the talented NIST team and the broader AI community to advance our scientific understanding and foster AI safety,” Kelly said. “While our first priority will be executing the tasks assigned to NIST in President Biden’s executive order, I look forward to building the Institute as a long-term asset for the country and the world.”
With its high-profile leadership now in place, the Biden administration wants the AI Safety Institute to play a key part in crafting and disseminating federal rules and regulations for how AI models and tools are created and monitored. One key mandate for the institute is to establish so-called red team testing protocols for major AI systems by July of this year. These safety tests will aim to ensure consumer and business protection prior to deployment. While the government has started requiring testing, universal standards are still lacking. The institute’s planned guidelines will fill this gap and boost trust in AI via more rigorous evaluations.
“The USAISI will advance American leadership globally in responsible AI innovations that will make our lives better. We must have a firm understanding of the technology, its current and emerging capabilities, and limitations. NIST is taking the lead to create the science, practice, and policy of AI safety and trustworthiness,” Tabassi said. “I am thrilled to be part of this remarkable team, leading the effort to develop science-based, and empirically backed guidelines and standards for AI measurement and policy.”
The AISI is a central pillar of the White House’s plans for AI regulation, but not the only facet. It also directed agencies across the executive branch to create standards for this testing, addressing related cybersecurity, biological, and other threats. The order pointed to the need for a regulatory framework to realize AI’s promise and avoid misuse. That included plans to address potential exploitation by hackers, create safeguards against AI-fueled discrimination, and plan for the impact on workers and the labor market. The AISI and the other plans in motion mesh with the White House-sponsored agreement among many of the leading generative AI companies to make sure those issues are considered during the development and rollout of new AI models.
“For the United States to lead the world in the development of safe, responsible AI, we need the brightest minds at the helm. I’m proud to announce that Elizabeth Kelly will lead the AI Safety Institute. I’m also thrilled that Elham Tabassi will expand her work at NIST and play a central role on our executive leadership team,” Secretary Raimondo said in a statement. “Together, they will provide the direction and expertise we need to mitigate the risks that come with the development of this generation-defining technology, so that we can harness its potential.”