Biden makes clear that AI cannot deliver nuclear weapons as he seeks to harness the power of new technology

() – He US President Joe Biden ordered its national security agencies on Thursday to take advantage of powerful new artificial intelligence (AI) technology in an attempt to compete with rivals such as China, while applying security measures to prevent its use for undemocratic purposes.

The new national security memorandum issued in the final months of Biden’s term seeks to strike a balance between unleashing the powerful potential of AI and protecting against some of its fearsome possibilities.

For example, the document prohibits the use of AI to circumvent existing protocols regarding the deployment of nuclear weapons and the use of technology to limit freedom of expression. More broadly, it prohibits agencies from using AI in ways that “do not align with democratic values.”

Those parameters may seem like obvious rules, but Biden and his team believe that setting them explicitly can help agencies better harness the power of tools they fear could be quickly used by global adversaries.

“Without clarity in policies and laws about what can and cannot be done, we are likely to see less experimentation and less adoption than if there were a clear path to its use,” said a senior government official.

In a series of discussions with his national security team dating back months, Biden sought to fine-tune the document to ensure its restrictions would apply even in times of crisis, the officials said.

Unveiling the new parameters during a speech at the National War College, Biden’s national security adviser, Jake Sullivan, framed the risks of managing the rise of AI.

“We have to get this right, because there is probably no other technology that is more critical to our national security and the years to come,” Sullivan said.

In the memo, the president also directed agencies to help U.S. companies protect AI technology from foreign spies and to continue advancing the production of semiconductor chips that are needed for many AI products.

The order sets deadlines for agency reviews that will expire long after Biden has left office. It’s unclear how former President Donald Trump would address the issue if elected. Vice President Kamala Harris has been active in the current administration’s AI efforts, including attending a summit on the topic in London.

Biden administration officials believe the rapid development of AI tools will trigger new competition among world powers. While largely developed in the United States, countries like China are rapidly applying the technology in military and civilian settings.

The United States has held talks with Beijing on this issue, but hopes to delve deeper into the riskiest applications of AI.

“We know that China is building its own technological ecosystem with a digital infrastructure that will not protect sensitive data, that can enable mass surveillance and censorship, that can spread misinformation, and that can make countries vulnerable to coercion,” he said. Sullivan.

“Therefore, we have to compete to offer a more attractive path, ideally before countries get too far down an unreliable path from which it can be costly and difficult to return,” he said.