The Biden administration on Monday issued sweeping rules governing how ai chips and models can be shared with foreign countries, in a bid to establish a global framework that will guide how artificial intelligence spreads around the world in the coming years.
With the power of ai growing rapidly, the Biden administration said the rules were necessary to keep a transformative technology under the control of the United States and its allies, and out of the reach of adversaries that could use it to augment their militaries, carry out cyberattacks and otherwise threaten the United States.
The rules impose several limitations on the number of ai chips that companies can ship to different countries, essentially dividing the world into three categories. The United States and 18 of its closest partners (including Britain, Canada, Germany, Japan, South Korea and Taiwan) are exempt from any restrictions and can freely purchase ai chips.
Countries already subject to US arms embargoes, such as China and Russia, will continue to face a previously existing ban on the purchase of ai chips.
All other nations (most of the world) will be subject to limits that restrict the number of ai chips that can be imported, although countries and companies can increase that number by entering into special agreements with the US government.
The rules are aimed at preventing China from obtaining the technology it needs to produce artificial intelligence from other countries, after the United States banned such sales to China in recent years.
But the regulations also have broader goals: to make allied countries the place of choice for companies to build the world's largest data centers, in an effort to keep the most advanced ai models within the borders of the United States. and his partners.
Governments around the world, particularly in the Middle East, have been pumping money into attracting and building huge new data centres, in a bid to become the next hub for ai development.
Jake Sullivan, President Biden's national security adviser, told reporters on Sunday that the rule would ensure that the infrastructure to train the most advanced artificial intelligence is in the United States or in the jurisdiction of close allies, and “that capability is not offshored.” “. like chips and batteries and other industries in which we have had to invest hundreds of billions of dollars to recover them.”
“The rule provides greater clarity to our international partners and industry, and counters the serious circumvention and related national security risks posed by countries of interest and malicious actors who may attempt to use advanced U.S. technologies against us,” he said. Sullivan. .
It will be up to the Trump administration to decide whether to keep the new rules or how to enforce them. In a call with reporters on Sunday, Biden administration officials said they had been consulting with the incoming administration about the rules.
Although Chinese companies have begun developing their own ai chips, the global market for such semiconductors is dominated by American companies, particularly Nvidia. That dominance has given the US government the ability to regulate the flow of artificial intelligence technology around the world by restricting exports by US companies.
Companies have protested those limitations, saying the restrictions could hamper harmless or even beneficial types of computing, anger U.S. allies and ultimately push global buyers to buy non-U.S. products, such as those made by China.
In a statement, Ned Finkle, Nvidia's vice president of government affairs, called the rule “unprecedented and misguided” and said it “threatens to derail innovation and economic growth around the world.”
“Rather than mitigating any threats, Biden's new rules would only weaken America's global competitiveness, undermining the innovation that has kept America ahead,” he said.
Brad Smith, president of Microsoft, said in a statement that the company was confident it could “fully comply with the high security standards of this rule and meet the technological needs of the countries and customers around the world who rely on us.”
The rules, which are more than 200 pages long, also establish a system in which companies that operate data centers, such as Microsoft and Google, can apply for special government accreditations.
In exchange for following certain security standards, these companies can market ai chips more freely around the world. Companies will still have to agree to keep 75 percent of their total ai computing power within the United States or allied countries, and not locate more than 7 percent of their computing power in any other nation.
The rules also establish the first controls on ai model weights, the unique parameters of each model that determine how the artificial intelligence makes its predictions. Companies establishing data centers overseas will need to adopt security standards to protect this intellectual property and prevent adversaries from gaining access to it.
artificial intelligence is rapidly transforming the way scientists conduct research, how companies assign tasks among their employees, and how the military operates. While ai has many beneficial uses, U.S. officials are increasingly concerned that it could enable the development of new weapons, help countries surveil dissidents, and otherwise upset the global balance of power.
Jimmy Goodrich, senior adviser for technology analysis at the RAND Corporation, said the rules would create a framework to protect U.S. security interests while allowing companies to compete abroad. “They are also forward-thinking and trying to preserve supply chains led by the United States and its allies before they are offshored to the highest subsidy bidder,” he said.
