US President Joe Biden on Monday issued a landmark executive order that aims to manage the risk of artificial intelligence while boosting its vast potential productivity upside.
The Executive Order on Safe, Secure and Trustworthy Artificial Intelligence includes clear guidance and resourcing to maintain US leadership in AI on national security grounds, and ties AI development to the disclosure and safety requirements of the Defense Production Act.
The executive order is a sweeping recognition by the US that if left unregulated, AI poses a threat to US national security and social safety and includes a comprehensive strategy for “responsible innovation”.
It establishes new standards for AI safety and security, privacy protections, advances equity and civil rights, sets up benchmarks for consumers and workers, promotes innovation and competition, measures to advance US leadership around the world.
The safety mechanisms are similar the approach taken in relation to the regulation of the nuclear industry, or biochemicals.
It mandates that developers of the most powerful AI systems must share their safety test results and other critical information with the US government.
“In accordance with the Defense Production Act, the Order will require that companies developing any foundation model that poses a serious risk to national security, national economic security, or national public health and safety must notify the federal government when training the model, and must share the results of all red-team safety tests,” the executive order says.
“These measures will ensure AI systems are safe, secure, and trustworthy before companies make them public.”
The US National Institute of Standards and Technology has been ordered to “set rigorous standards for extensive red-team testing” to ensure safety before the public release of an AI. The Homeland Security department would apply those standards to critical infrastructure sectors and establish an AI Safety and Security Board, the executive order says.
The Departments of Energy and Homeland Security have also been ordered to address AI threats to critical infrastructure, as well as chemical, biological, radiological, nuclear, and cybersecurity risks.
President Biden’s AI executive order was not unexpected, and has been issued just days ahead of a global AI safety summit being convened by British Prime Minister Rishi Sunak on Wednesday and Thursday at the Bletchley Park, the famed home of the code-breakers in the second world war.
US vice-president Kamala Harris is scheduled to attend the AI Safety Summit and is expected to provide an update on the measures announced by President Biden on Monday.
Deputy Prime Minister Richard Marles and Industry minister Ed Husic will attend the summit on behalf of the Australian government. Mr Husic will participate in ministerial level discussions on day one of the summit, while Mr Marles will represent the Prime Minister at the leaders’ discussions.
At the summit, Mr Husic will highlight work already underway in Australia, and discuss possible guardrails to support AI transparency, accountability and fairness. He will also meet with counterparts to explore collaboration opportunities emerging technologies like quantum and robotics.
“The UK AI Summit is important and timely. It brings leaders from around the world, and across government, academia and industry, to explore solutions to better ensure the safe and responsible use of AI,” Mr Husic said in a statement ahead of the summit.
“In Australia, we recognise that AI and automation can provide a massive opportunity to boost productivity, but this will only be possible if the right guardrails are in place to build public trust,” he said.
Do you know more? Contact James Riley via Email.