US issues new rules on use of AI by security establishment

0

Unlock the US Election Countdown newsletter for free

US President Joe Biden has announced new rules governing the use of artificial intelligence that will bar the Pentagon and intelligence communities from using the technology in ways that do not “align with democratic values”.

Biden will publish the new guidelines in a national security memorandum he is due to sign on Thursday. It is the first directive outlining how the US national security apparatus should use AI and aims to set an example for other governments looking to use and expand the technology responsibly, officials said.

They added that the new rules were designed to encourage the use of and experimentation with AI, while ensuring that government agencies do not employ it for activities that could, for instance, violate the right to free speech or sidestep controls on nuclear weapons.

“Our memorandum directs the first-ever government wide framework on our AI risk management commitments . . . like refraining from uses that depart from our nation’s core values, avoiding harmful bias and discrimination, maximising accountability, ensuring effective and appropriate human oversight,” US national security adviser Jake Sullivan said in a speech on Thursday morning.

The guidelines are not legally binding and if he wins next month’s presidential election, Donald Trump could choose not to enact them. Vice-president Kamala Harris has played a key role in shaping the Biden administration’s efforts on AI and is expected to focus on emerging technologies if she is elected.

The directive is the latest effort by the Biden administration to try to foster use of AI as the US seeks to compete with China, while responding to concerns about the potential misuse of the technology.

The rules focus on the national security applications of AI, such as in cyber security, counter-intelligence, and logistics and other activities that support military operations.

The US has undertaken a number of measures in an effort to maintain a strategic advantage on the technology, including issuing export controls aimed at slowing China’s development of advanced AI.

Biden last year signed a sweeping executive order that compelled private companies, whose AI models could threaten US national security, to share safety information with the US government.

The new memorandum directs the US intelligence community to prioritise collecting information on competitors’ AI activities. It also designates the AI Safety Institute in Washington as responsible for inspecting AI tools to prevent their misuse before they are released.

#issues #rules #security #establishment

Leave a Reply

Your email address will not be published. Required fields are marked *