White House: We're "clear-eyed" about weaponizing A.I.

The U.S. could be a world leader in establishing a code of conduct regarding the use of militarized artificial intelligence, the White House chief of staff said.
Written by Stephanie Condon, Senior Writer

Paying heed to the artificial intelligence researchers who have issued dire warnings about an inevitable A.I. arms race, a White House official on Tuesday said the U.S. Defense Department is taking a "clear-eyed" approach to using A.I. and machine learning.

"We are looking through the DoD at how we can use A.I., machine inearning, to increase our defenses," White House chief of staff Denis McDonough said in an online chat regarding automation. He was joined by transportation entrepreneur and author Robin Chase, as well as author and futurist Martin Ford.

At the same time, McDonough said, the administration is "also very clear-eyed in making sure there's appropriate human interaction in any kind of weaponization of any of this artificial intelligence."

The U.S., he said, is aiming to be a world leader "in establishing a code of conduct and a set of understandings" in the field, "including, perhaps, arms through kinds of control arrangements or negotiations that take some of this stuff off the table."

Returning to an earlier portion of the online discussion regarding the economic impact of automation, Chase argued that "income inequality will generate a larger security threat than those other things. It's already a huge issue, and if automation makes it worse, it will be even more acute."

Both Chase and Ford over the course of the discussion endorsed the idea of universal income to compensate for job losses due to automation and higher productivity.

Earlier in the conversation, Ford also responded to McDonough observations that machine learning and A.I. can be used to detect and defend systems against cyber attacks.

"In areas like security, I think it will be essential," Ford said of A.I. "There's an arms race happening there. The criminals, the people on the other side, are using advanced technology to attack our systems. Many people would agree the only way we're going have security is if we can leverage A.I. on our side to protect against that."

Editorial standards