‘Zero Trust’ Approach to AI Will Continue
The National Telecommunications and Information Administration will continue its “zero trust” approach, which assumes any hardware or software technology cannot be trusted, when it comes to grappling with issues surrounding artificial intelligence, according to the agency’s top official today.
“AI is not the future … it’s already in use in the federal government in multiple applications,” said Evelyn Remaley, acting assistant secretary of commerce for communications and information at the NTIA, an executive branch agency that advises the president on telecommunications and information policy matters.
Along with two other NTIA officials, Remaley spoke during an event hosted by the Federal Communications Bar Association entitled, “Artificial Intelligence, Machine Learning, and Robotics Committee Lunch and Learn,” during which she was asked how the NTIA and the new administration will move forward with the advancement of emerging technologies, particularly AI, while addressing issues of cybersecurity and ethics.
However, Remaley explained, the NTIA is not a regulator, but a “player pulling all the pieces together” between the White House and its National Security Committee. It also works with many other agencies in advancing the administration’s AI research efforts.
Each agency will then “dive deeper” into the specific applications of this tech in their organization, she added. For example, the National Oceanic and Atmospheric Administration, will look at how AI can impact weather. As the NTIA falls under the Department of Commerce, she explained, it looks at AI’s application within commerce and competition.
The zero-trust approach, Remaley explains, “builds in controls from the bottom up” to ensure that security risks are taken into account at the basic level for all federal agencies.
“The conversations between cybersecurity and AI are obviously intrinsically linked to every issue that we work on,” said Jaisha Wray, associate administrator for international affairs at the NTIA. Wray works with international partnerships like the Organization for Economic Co-operation and Development and foreign allies on how to best “protect” and “promote” the use of this technology.
AI is like a “thread woven into every issue” in the global digital economy from national security and data privacy concerns to algorithmic transparency, accountability, and ethical issues like the spread of harmful extremist content or disinformation, she added.
One of the top concerns of the U.S. and its allies is the abuse of AI technology on behalf of authoritarian governments to surveil their people – going against basic civil liberties, Wray said.
In 2019, the OECD published a set of recommendations on AI principles seeking to “harness [its] power,” Wray said, marking the first international agreement to promote a “trustworthy” AI tech aligned with “human rights and democratic values.” It also instructed policymakers to form their own national policy. International cooperation and collaboration are essential to fighting the “spread of abuse,” she added.
In the midst of a pandemic that has made “[everyone do] everything online,” Remaley said these emerging technologies have become “so prevalent” in everyday life that these public policy issues – equity, protection from “malign influence,” privacy – will definitely be taken up by President Biden’s administration. All panelists agreed but said it was too soon to tell what the new cabinet’s AI priorities will be.
The U.S. must continue to find a way to “manage public policy” while enjoying the advantages of AI, Remaley said, which can even help the country deal with current global issues “from climate change to public health.”