Trump signs executive order to boost AI technology – February 11, 2019
President Trump on Monday signed an executive order laying out a national plan to boost artificial intelligence (AI) technology, amid growing concern that the U.S. is losing out to China.
The executive order directs federal agencies to prioritize and set aside funding for AI programs, while opening up the way for researchers and developers to access more government data.
“Continued American leadership in Artificial Intelligence is of paramount importance to maintaining the economic and national security of the United States,” Trump said in a statement, accompanying the order.
The order directs all federal agencies to look into launching and expanding AI initiatives that promote their missions.
The order also asks a coalition of government bodies to develop a set of national “regulatory” standards around AI, which the U.S. currently lacks. Administration officials over the weekend told reporters that the standards will address some of the ethical issues posed by AI, including privacy concerns around increased access to government data on U.S. citizens.
Finally, the executive order directs the National Council for the American Worker and AI select committee to set up “fellowship and training programs” to help U.S. workers learn the skills needed to work with and develop AI technologies.
“As the pace of AI innovation increases around the world, we cannot sit idly by and presume that our leadership is guaranteed,” the White House’s Office of Science and Technology Policywrote in a fact sheet on the order.
The order is light on details but ambitious in its mission.
Michael Kratsios, the deputy assistant to the president for technology policy, in a statement said the executive order aims to prioritize AI, “preparing America’s workforce for jobs of today and tomorrow.”
“Artificial intelligence is something that touches every aspect of peoples’ lives, from the way they receive medical care to the way they grow or eat their food to the way that energy is extracted or resources are extracted from across the country,” a senior administration official told reporters over the weekend. “What this initiative attempts to do is bring all of these together.”
For example, the National Oceanic and Atmospheric Administration could use satellite data to improve severe weather predictions and the Department of Transportation could dedicate more resources to automated vehicle safety.
But the order does not lay out a specific timeline or tangible goals on AI, experts told The Hill. It also does not direct Congress to appropriate any new funds for AI programs, instead asking federal agencies to take the lead in setting aside more money and resources.
Administration officials said a more detailed plan would come over the next six months.
The order comes two years after China announced a detailed plan to become a global leader in artificial intelligence. Other countries, including South Korea, Canada and France, have also made major commitments to improving their AI technologies.
Industry leaders and tech experts have been calling on the Trump administration to step in and prioritize AI to stave off the advances from China and other competitors.
“I think that the U.S. is still the dominant force [on AI] and I think that China is rapidly catching up,” Oren Etzioni, the CEO of the Allen Institute for Artificial Intelligence, told The Hill.
Etzioni said he worries the executive order is “too little, too late,” but added it will be hard to tell until the mandates are implemented.
The Obama administration in 2016 released a detailed plan on artificial intelligence, which was abandoned under the Trump administration.
Ryan Calo, a cyber law expert at the University of Washington, told The Hill that Trump’s executive order appears to incorporate some elements of former President Obama’s AI plan.
Lawmakers also praised the move.
“Artificial Intelligence has great potential to benefit the American people, while enhancing our nation’s security and growing our economy,” said Senate Commerce Committee Chairman Roger Wicker (R-Miss.) in a statement. “Today’s executive order will ensure that the United States remains a leader in emerging technologies and scientific development.”
Tech companies see an opportunity to cash in on AI. A number of prominent firms released statements on Monday calling the order a “step” toward a comprehensive national strategy on AI.
“Today’s order is a critical step in the launch of America’s national AI strategy,” IBM said in a statement to The Hill. “We commend and welcome the order’s emphasis on specific priorities that IBM had recommended, such as the ethical advancement of AI, expanding 21st century apprenticeship opportunities to build an AI-ready workforce, leveraging government data to accelerate AI development that can deliver shared prosperity, and prioritizing AI in federal research and development.”
Intel in a statement said it makes “perfect sense” for federal agencies to play a “key role in AI implementation.”
Dell called on the administration to follow up on the order with “tangible goals and funding.”
Efforts to have tech companies work closely with the government though will face scrutiny. Some companies have faced criticism for such work, particularly on AI defense matters. Google last year passed on a $10 billion Pentagon contract following pushback from many employees.
Google CEO Sundar Pichai at the time said Google’s standards prevent it from developing AI that causes harm, facilitates weapons development, breaks international norms on surveillance or violates international human rights laws.
Google, Amazon and Microsoft did not respond to The Hill’s requests for comment Monday.
Some federal agencies have already donated significant resources toward AI research and development, most prominently the Department of Defense (DOD) and the Department of Transportation.
“This [order] is, as we see, an extrapolation of what’s happening within the DOD across the U.S. government,” Todd Probert, vice president of Raytheon Intelligence, told The Hill.
“There’s been a lot of interest and increased requirement across the defense and intelligence communities,” Josh Elliot, the director of Booz Allen Hamilton’s analytics and AI division, told The Hill. “The federal civilian agencies are following suit.”
“I think that there are opportunities for all aspects of government to employ these technologies but it’s important for each organization to take stock of their mission and understand what their goals are for employing machine learning,” Elliot added.
Most AI experts who spoke with The Hill said there are significant concerns the government should address, particularly with AI technologies that are biased against minorities and marginalized communities.
“This administration has a concerning track record of keeping its promises to uphold and protect privacy, civil rights, and consumer protection laws,” research center AI Now said in a statement to The Hill. “We are skeptical that the passing mention of these protections will result is any serious efforts to build in appropriate legal, ethical, and policy safeguards to ensure that AI systems are deployed responsibly.”
Calo, the cyber law expert, told The Hill there needs to be an examination of “the way in which artificial intelligence can be biased, the way it can disproportionately harm vulnerable populations.”
He said government needs to regulate how it procures AI technologies “so that we don’t ‘unleash AI’ on the world without thinking about its social impacts.”