US government proposes a 'light-touch' to developing AI regulation

The White House has proposed principles that seek to prevent federal agencies from "needlessly hamper[ing] artificial intelligence (AI) innovation and growth".

In a draft memorandum [PDF] released on Tuesday, the government presented its vision for how AI regulation should be developed, saying that federal agencies should have a "light-touch" and not have "such an impossibly high standard" when assessing AI innovation and growth and its effects on society.

"Where AI entails risk, agencies should consider the potential benefits and costs of employing AI, when compared to the systems AI has been designed to complement or replace," the government wrote.

The US government proposes 10 principles, ranging from building public trust in AI to ensuring civilians can participate in the AI rule-making process.

Among the principles is Risk Assessment and Management, wherein the US government wrote that it should not be necessary for federal agencies to mitigate every foreseeable risk. The US government attributed this belief to the "foundational principle that all regulatory policy involve trade-offs".

"Instead, a risk-based approach should be used to determine which risks are acceptable and which risks preset the possibility of unacceptable harm, or harm that has expected costs greater than expected benefits," the government wrote.

Following the theme of "limiting regulatory overreach", the US government is also pushing for flexible approaches to developing AI regulations as part of its principles. It envisions AI regulation that can be adapted to rapid changes as "rigid, design-based regulations … will in most cases be impractical and ineffective".

On the privacy front, the government also put forth two principles -- Disclosure and Transparency and Safety and Security -- that recommend for federal agencies to consider what is appropriate disclosure and transparency, as well as what are appropriate controls for preserving the confidentiality, integrity, and availability of information processed by AI systems.

At the same time, the US government has urged other nations to avoid heavy handed approaches to regulating AI as it wants to be the table-setter for how AI policies are shaped. 

"Governments elsewhere are co-opting companies and deploying their AI technology in the service of the surveillance state, where they monitor and imprison dissidents, activists and minorities, such as Beijing's treatment of the Muslim Uyghurs. The best way to counter this dystopian approach is to make sure America and our allies remain the top global hubs of AI innovation," chief technology officer of the United States Michael Kratsios wrote in an op-ed on Wednesday.

"Europe and our other international partners should adopt similar regulatory principles that embrace and shape innovation, and do so in a manner consistent with the principles we all hold dear." 

The European Commission's High-Level Expert Group on Artificial Intelligence issued a set of ethical guidelines last year. These guidelines are currently being considered by EU leaders. 

While the US has called for other nations to adopt its AI principles, the US itself has so far rejected offers to participate in various global AI initiatives, such as the Global Partnership on AI, despite it being backed by every other G7 nation. 

It has also issued bans that are aimed at keeping key technologies away from rival powers like China, having last week decided to make it harder for US companies to export certain types of AI software by requiring them to apply for a licence to send the tech overseas except when it is shipped to Canada.

The announcement of these 10 principles follows US President Donald Trump signing an executive order in February last year to fast track the development and regulation of AI in the United States. 

The executive order launched the American AI Initiative, which places United States resources toward ensuring that AI technology is made in the US. 

"The policy of the United States government [is] to sustain and enhance the scientific, technological, and economic leadership position of the United States in AI," Trump wrote at the time.

The White House has opened the AI principles up for a 90-day period of public consultation, which will then be followed by federal agencies having 180 days to devise plans for implementing these principles. 

"By working together, we will shape the policies that guide how AI is developed and deployed so that all people and communities can enjoy the benefits and opportunities it provides," Kratsios said.

Related Coverage

CES 2020: Citing "mystery drones," US Transportation Secretary advocates new rules

The mysterious drones flying over Colorado and Nebraska offer a "timely illustration" of why drones should be equipped with remote IDs, the administration official said.

Google DeepMind gamifies memory with its latest AI work

Google DeepMind scientists built a computer program that gives signals from future to past, in a kind of theoretical model that feels like things people do when they learn from their mistakes. Just remember, it's only a game.

2020 is when cybersecurity gets even weirder, so get ready

AI-powered deepfakes, ransomware, IoT, and 5G all mean that protecting your data is about to get a lot harder.

Tech trade ban: Exporting AI is going to get harder, says US

US imposes new trade restrictions on companies exporting new technologies to rival powers.

China aims to overtake US as global leader in AI innovation (TechRepublic)

The US still ranks far ahead of both China and the EU in AI innovation, but China is making moves to change things, according to a Center for Data Innovation report.