NSI Experts Weigh In: “Maintaining American Leadership in Artificial Intelligence” Executive Order

Today, President Trump signed an Executive Order on “Maintaining American Leadership in Artificial Intelligence”.  Our experts weighed in about what this means about America’s future stance towards Artificial Intelligence. 

February 11, 2019


Dmitri AlperovitchDmitri Alperovitch – NSI Visiting Fellow; Co-Founder/CTO, CrowdStrike

Today, AI is at the center of most major technological advances in areas as varied as cybersecurity, self-driving cars or development of cancer treatments. In cybersecurity, for example, these technological advances include enabling the defenders to recognize and stop never-before-seen attacks and being faster in detecting and responding to attacks.

It is vital for prosperity and security that the US keeps its leadership position in development and implementation of AI technologies. Government leadership and funding on this topic is crucial.

 


Lauren BedulaLauren Bedula – NSI Visiting Fellow; Vice President, Beacon Global Strategies

The Executive Order signals that Artificial Intelligence (AI) is a White House priority with the full backing of the President.  This has a galvanizing effect and elevates AI as a critical national priority.  The Department of Defense and Intelligence Community have recognized the significance of AI through initiatives such as the Office of the Director of National Intelligence’s Augmenting Intelligence Using Machines program, and DoD’s Joint AI Center.  The White House’s push to accelerate national leadership on AI will further stimulate the federal government’s civilian agencies’ efforts, and enhance the U.S. government’s posture on coordinated research and development.  The timing of the EO also aligns well with the launch of the National Security Commission on AI, created by the National Defense Authorization Act for fiscal 2019.

Access to and management of data will continue to slow adoption of AI technologies, and migration to the cloud will be key to these efforts.  As the U.S. government organizes around the future of AI, the private sector should be consulted as an expert and partner in both technology and workforce development.


Bryson Bort – NSI Fellow; Founder & CEO, SCYTHE

The Administration actually does agree with their Intelligence Chiefs on something: Artificial Intelligence. In January, ODNI published the National Intelligence Strategy highlighting both the threat, promise, and need for investment in AI. The primary threat is China who through a combination of outright theft/espionage and state-directed investment to lure foreign scholars has positioned itself to strategically challenge US innovation dominance by 2025. The Administration builds upon these concerns by laying out a (high-level) plan of R&D investment and engagement that vaguely directs agencies to prioritize AI now and in 2020 funding. Contrast that lack of clear funding with just a few examples:

 

China – $5B fund to grow a $150B industry
France – €1.5B investment
Canada – the first government release a formal AI strategy with C$125M

Priorities mean resources. Or, are we waiting for the computers to tell us how much to invest?


Zach Graves

Zach Graves – NSI Visiting Fellow; Head of Policy, Lincoln Network

“The President’s Executive Order on ‘Accelerating America’s Leadership in AI’ is a monumental action that ties together both public and private sector efforts around artificial intelligence.  This includes calling for the development of several strategic plans around standard setting, research spending, and AI governance issues which will help improve coordination and strategic planning in the federal government.  Of potential concern is Sec. 8’s call for the creation of an action plan to ‘protect the US advantage in AI… against strategic competitors and adversarial nations.’  While there are legitimate national security concerns around AI, these must not be used as an excuse for premature heavy-handed regulation.  For example, the Commerce Department has already been exploring export controls on AI technology – a failed policy approach that was used in the 90s on encryption, and which could put the brakes on AI rather than accelerate its development. “


Kristen HajdukKristen Hajduk – NSI Visiting Fellow; Regional Director for the National Capital Region, MD5 – The National Security Technology Accelerator

“The Executive Order on ‘Maintaining American Leadership in Artificial Intelligence’ takes one step forward in guiding our country towards a vibrant, innovative, and ethical support in the development of AI.  The EO rightly acknowledges that effective utilization of AI requires a community of innovators who are given the tools to support the future public and private sector needs of this growing industry.  In order for the USG to effectively foster and tap into private sector innovation, it will be necessary and important for all USG organizations to rethink and retool acquisitions processes that may be burdensome or prohibitive for burgeoning startup and venture hubs across the country.  The Defense innovation ecosystem has led to several important efforts to support the EO, including the Joint Artificial Intelligence Center and the National Security Commission on Artificial Intelligence.  By working together, the USG and US private industry will continue to push the leading edge of technological advancement.”


E. Grant Haver – NSI Policy Program Coordinator

President Trump’s ‘Maintaining American Leadership in Artificial Intelligence’ Executive Order is far from the ‘[w]e choose to go to the Moon’ approach which is necessary in this area.  He had the prime opportunity to roll out the American Artificial Intelligence Initiative during his State of the Union address, but he chose not to.  America needs a whole of government and potentially a whole of society approach to dealing with the ethical, legal, and military questions surrounding AI and the president does not seem to be going all in.

One area that this Executive Order addresses is the need for the government to attract and train personnel on the latest advances in AI.  However, it fails to address the issue of retention which is difficult in the best of times because skills in this area are so valuable in the private sector.  These issues are made worse when those who choose to take a pay cut to serve their country are not sure whether they will be getting their next paycheck.

Although this Executive Order is far from sufficient to actually maintain American leadership in Artificial Intelligence, it makes nods in the right direction especially around data transparency.  AI is powered by large sets of training data and the government is sitting on a wealth of it.  If the government can learn how to effectively partner with the private sector to give access to this data while protecting citizens privacy and civil liberties, then the government, industry, and ultimately the American people will benefit.


Jamil N. Jaffer – NSI Founder; former Chief Counsel and Senior Advisor for the Senate Foreign Relations Committee and former Associate Counsel to President George W. Bush

“The artificial intelligence Executive Order signed today by the President provides a strong starting point for expanding our government’s commitment to our nation investing and leading in AI.  The problem is that we are already behind in this race, with China having committed to creating a $1 trillion AI economy in the next decade.  We cannot afford more government examinations of its authorities and priorities; rather, we need specific, detailed plans for providing significant, sustained government funding for basic research and advanced national security applications, as well as the creation of incentives for further private sector investment and leverage.  So, while the new EO is certainly a step in the right direction, it is not nearly enough.  The President and his team need to stop focusing on small ball issues that keep taking us to the brink of a government shutdown and instead partner with Congress to get something definitive done on AI in the near future.


Geof Kahn  – NSI Visiting Fellow; former Senior Advisor to the Chief Operating Officer at the Central Intelligence Agency

The Executive Order is an encouraging signal that the White House is using its budget, policy, and convening authority to implement a national AI strategy to empower American competitiveness and innovation.  That strategy smartly identifies much needed funding for R&D that will benefit mainstream American industries, while highlighting the jobs, ethics, and privacy implications that come with our rapidly increasing adoption of AI tools.  It will be important to see how this EO is reflected in the President’s FY20 Budget Request in March and how the Administration pursues these initiatives in partnership with the private sector.


Andy KeiserAndy Keiser – NSI Fellow; former Senior Advisor, House Intelligence Committee

In launching the American Artificial Intelligence (AI) Initiative, President Trump signals the right executive emphasis on the type of next-generation technologies that will propel the U.S. economy into the future like AI.

The Order also touches on the critical components of success in AI, including: focus of effort across government, industry and academia; development of standards; promoting technical and math and science education; protecting civil liberties and promoting AI in the right places internationally.

As with any policy, the devil will be in the details as the Order is implemented and federal dollars are appropriated in Fiscal Year 2020 and beyond.


Andrea LimbagoDr. Andrea Little Limbago – NSI Senior Fellow; Chief Social Scientist, Virtru

The U.S. has some catching up to do to simultaneously shape the ethical parameters of AI, including privacy and security implications, while promoting innovation. Today’s Executive Order on AI is a welcome first step at prioritizing AI  but is long overdue and lacks any reference to countering advances in adversarial AI. While policy generally does lag behind technology, over the past two years well over a dozen countries introduced national initiatives to promote AI. The U.S. has been noticeably absent from this global discussion. The Executive Order appropriately describes AI as a national security imperative that requires strong U.S. leadership to ensure AI benefits society without compromising American values or privacy. In contrast, authoritarian regimes are quickly progressing AI frameworks that build upon machine learning for offensive attacks such as compromising data, diffusing disinformation, or other interference operations. Given this contrast, the Executive Order should have more specifically addressed the ethical issues, biases, and attacks that can emerge through adversarial AI or abuse. While the Executive Order highlights the potential for benefits, any national strategy would be myopic and detrimental to national security if the the potential for AI misuse and abuse is omitted. Nevertheless, the Executive Order is an encouraging sign that the U.S. government will increasingly prioritize AI and take the steps to incentivize AI investments, grow the workforce to support U.S. dominance in AI, and assert U.S. leadership through international engagement. 


Harold Moss – NSI Visiting Fellow; Senior Director Strategy & Business Development, Web Products, Akamai Technologies

Artificial Intelligence has penetrated the lives of millions of Americans through connected home devices that influence as well as provide access to personal information. It is essential that America establishes the skills necessary to compete in this space as well as define the working parameters for which this technology is applied.

According to research by IoT News, as may as 90% of the US population have made purchases of a home connected device, and nearly 70% already have a voice controlled system such as Amazon Alexa or Google Home. As artificial intelligence becomes an embedded technology in our personal lives, it offers not only new methods of connecting but also new means of gather information or influencing our population. The recent Executive Order highlights the importance of American leadership in the space, and touches on the role of privacy, as well as the highlighting the need for addressing adoption barriers.

The Executive Order, while light on details is never the less laudable in that it addresses emerging technologies proactively. It will be essential that the security implications are the initial focus, as the path to failure will most certainly fall in the areas of confidence and privacy. It will be imperative that legislators with their private sectors counterparts  take up the mantle of defining meaningful guidelines quickly and comprehensively in Order to decrease threats to our population, business innovation, and our future democracy.


Bryan Smith – NSI Senior Fellow; former Budget Director, House Permanent Select Committee on Intelligence

The President’s Executive Order on Artificial Intelligence is an impressive ‘plan for the plan’ (or to be precise, a plan for several related plans).  This is a thoughtful and well-constructed architecture for all the work the federal government needs to accomplish, in close collaboration with its many partners, domestically and world-wide.  The E.O. underscores the critical importance and wide-sweeping nature of artificial intelligence, which is already permeating every business vertical and horizontal, and even our every-day lives.  A.I. promises to be just as transformational for national security.

Even so, the seminal developments in A.I. will continue to come from gifted mathematicians and scientists in universities and technology companies in every corner of the globe, supported by investors with foresight, patience, and high tolerance for risk.  There is a parallel here with the recent transformation of oil and gas production through hydro-fracture and horizontal drilling.  American politicians routinely called for American energy independence, but this dream could only become reality through the genius of a modest number of technical innovators, backed by private investors willing to bet the house on them.


Glenn Sulmasy – NSI Visiting Fellow; Provost and Chief Academic Officer, Bryant University

President Trump’s Executive Order today – the American AI Initiative (AAII) – is welcome news to many within national security circles, as well as academics and administrators in higher education.  We need bold policy statements, such as this, to be the catalyst for Americans to recognize the importance AI will play in ALL of our lives within the next two-to-three years, and beyond.  This ‘call to action’ should be viewed as analogous to Kennedy’s charge for Americans to be the first to put a ‘man on the moon.’

Artificial intelligence is coming at us quicker than workplace revolutions of the past.  This so called 4th Industrial Revolution is one that will continue to evolve and will be incredibly disruptive to our workplace.  Beyond the call to research and development, there needs to be a consciousness of what AI will do the workplace within American society.  While essential for the AAII to emphasize R&D, the document goes further to prioritize fellowship and training programs to help American workers gain AI-relevant skills.   The beauty of the document is that it demonstrates the will of the government, and the American need, to place appropriate emphasis on AI, it goes further to recognize and prepare for the disruptive nature of the technology that will accompany the new armies of robots.

Within higher education, it is critical to get ahead of the coming robot revolution.  The AAII must continue to support, not just graduate programs, but undergraduate programs that are dedicated to Artificial Intelligence.  Universities must adapt quickly to the changing environment to ensure the students matriculating have at least some basic competencies within the AI domain.   This baseline knowledge will be required by all students graduating by 2022.  As such, my hope is the administration will partner with universities to ensure learning outcomes within all disciplines soon require some form of ‘data literacy.’  To do otherwise, at this juncture in history, would be for academicians to commit malpractice.

While we appropriately race to compete with the EU and China (who are clearly dedicated to making major advances in AI before the US to gain myriad advantages), it remains important to reflect upon the words of the internationally known late genius Stephen Hawking’s warning, ‘The genie is out of the bottle. We need to move forward on artificial intelligence development but we also need to be mindful of its very real dangers. I fear that AI may replace humans altogether. If people design computer viruses, someone will design AI that replicates itself. This will be a new form of life that will outperform humans.’


Disclaimer: The views and opinions expressed in this analysis are those of the authors and do not necessarily reflect the official policy or position of the National Security Institute or any agency of the U.S. government. Assumptions made within the analysis are not reflective of the position of the National Security Institute or any U.S. government entity.