Laurent - stock.adobe.com
Editor's note: This story was updated on Jan. 12, 2021
On Jan. 12, days before President-elect Joseph Biden is scheduled to be sworn into office, the White House Office of Science and Technology Policy under the Trump administration established the National Artificial Intelligence Initiative Office.
The office, according to a release from the administration, will oversee and carry out U.S. AI strategy, as well as serve as a national hub for lawmakers and government employees, and those in the private sector and academia, to collaborate on AI research and policymaking.
The creation of the office was required by the recently passed National Artificial Intelligence Initiative Act of 2020, a broad new law that seeks to increase government and institutional funding for AI research, as well as boost collaboration on AI between government agencies and the private sector. It’s unclear how the Trump administration’s AI strategy will align with the incoming Biden administration.
Draft guidelines preceded the new office
In a memorandum released on Jan. 7, 2020, the White House presented a series of guiding principles that organizations should consider when drafting AI regulations, while making it clear that any adopted regulations should not hamper AI innovation.
The memo highlights 10 policy considerations that should guide government agencies in the development and deployment of AI regulations. It notes that the use of AI by the federal government falls outside the scope of the memo.
The document shows that the White House understands AI will have a huge impact socially and economically, and that the government is attempting to embrace the reality of AI, according to an analyst.
AI is "simply set of technology tools, what you have to regulate is the use of those tools," said Alan Pelz-Sharpe, founder of Deep Analysis.
Regulating AI is considerably harder than defining telecom standards such as CDMA (code division multiple access) or GSM (global system for mobile communication), he said.
"Over time, there will have to be consideration of tax and employment impacts from AI as jobs are automated. Similarly, there will have to be consideration of the ownership and responsibility of decisions made by AI," he continued. "The goal is to start somewhere, and though lax, [the memo] does achieve that."
The guiding principles for agencies crafting AI regulations include: public trust in AI, public participation, scientific integrity and information quality, risk assessment and management, flexibility, fairness and non-discrimination, disclosure and transparency, safety and security, and interagency coordination.
Alan Pelz-SharpeFounder, Deep Analysis
The principles, however, while looking good in theory appear to lack bite as the memo comes with a caveat: regulations should not discourage innovation or growth.
"Promoting innovation and growth of AI is a high priority of the United States government. Fostering innovation and growth through forbearing from new regulations may be appropriate," the memo says. "Agencies should consider new regulation only after they have reached the decision, in light of the foregoing section and other considerations, that Federal regulation is necessary."
In a sign that the government wants to show the tech industry it is working on AI, Michael Kratsios, the chief technology officer of the United States, spoke on the new principles at CES 2020 in Las Vegas, Jan. 8. The sprawling conference and exhibition is the country's top consumer electronics and technology show.
Necessary AI regulations?
Just how much oversight the AI industry needs has long been a point of contention among AI vendors, analysts and activists. While many have spoken forcefully about the dangers of AI and the need for strong regulations, many others have said that too many regulations will stifle growth, as the White House contends. It's a complex issue.
"I would argue we need clear-cut and fairly strict AI regulations to stimulate enterprise adoption. The fog around what is and isn't AI, what is safe, what is not, what is ethical and what is not, is stifling adoption today," Pelz-Sharpe said.
The draft memo came a day after new U.S. regulations restricting the sale of geospatial AI software in foreign countries took effect. The interim restrictions on the sale of a technology that has vast military potential are likely aimed at countries such as China.
In early 2018, President Trump signed an executive order directing federal agencies to increase spending on AI regulation, training and research. Called the American AI Initiative, the order called on U.S. agencies to set standards on AI development and use by businesses.