The United States' top intelligence agency has appointed a research official to spearhead the intelligence community's efforts on AI.
John Baylor, chief science and technology advisor to Director of National Intelligence Avril Haines, has been named chief artificial intelligence officer in the Office of the Director of National Intelligence. Mr. Byler acknowledged his additional role today in a speech at an event hosted by the Intelligence and National Security Alliance in Arlington, Virginia.
Mr. Byler currently leads the Council of Chief AI Officers, which is comprised of 18 departments in the intelligence community, including the CIA, the National Security Agency, and the Defense Intelligence Agency. He said the council, which reports to Haynes, has been meeting every two weeks for the past two months.
“Our focus as a group is AI governance,” Beiler said.
He said the group is developing the first IC-wide directive on AI. Learn what intelligence agencies need to do to implement AI and machine learning.
“Documents, standards, etc. [application programing interfaces]”What kind of data documentation is needed, how all of this fits together, responsible hiring, ongoing monitoring,” Baylor said, explaining what the directive entails. “The responsibility of individual developers, the responsibility of managers and leaders. We are focused on responsible and ethical adoption.”
He added that the directive would also provide for civil liberties and privacy protections that must be included in algorithms developed by intelligence agencies.
The new AI Council is also leading an update to ODNI’s AI strategy.
“We want to make sure we have one unified view of what we think is important for AI and IC to drive the conversation around resources,” Beiler said. .
Concerned about China and other countries' rapid advances in AI, lawmakers are calling on intelligence agencies to prioritize the adoption of AI with safeguards.
The National Defense Authorization Act of 2024 would establish new policies for “the acquisition, introduction, development, use, coordination, and maintenance of artificial intelligence capabilities,” including minimum guidelines for the performance of AI models used by spy agencies. Instructs DNI.
Mr. Byler has a background in data science and machine learning. Prior to joining ODNI in 2019, he led research programs on human language technologies, machine learning, and AI vulnerabilities at the Intelligence Advanced Research Projects Agency.
At ODNI, he also helped lead the Augmentation of the Intelligence Community with Machines, or “AIM” strategy. With many intelligence agencies processing large amounts of data, AIM's goal is to coordinate the deployment of AI and automation across spy agencies.
Spy agencies have been using forms of artificial intelligence and machine learning for decades, but the emergence of widely available large-scale language models like ChatGPT brings new considerations and new urgency to the AI race. Sex has been added.
“A lot of it is focused on helping the people who are using these tools understand them,” Beiler says.
ODNI already funds a variety of training and upskilling programs across the intelligence community. He acknowledged that generative AI and other large-scale language models have challenges, including illusory errors, copyright issues, and privacy concerns.
“We want a broad base of analysts, collectors, and IC employees to become familiar with these things and understand some of these failure modes, but in a way that doesn't immediately disable the technology. We will do that,” Byler said. “That’s the difficult part of upskilling the entire workforce.”
Copyright © 2024 Federal News Network. All rights reserved. This website is not directed to users within the European Economic Area.