Game Changers: Artificial Intelligence Part II; Artificial Intelligence and the Federal Government
Statement of Keith Nakasone
Deputy Assistant Commissioner, Acquisition Operations, Office of Information Technology Category (ITC), U.S. General Services Administration
Before the Subcommittee on Information Technology of the Committee on Oversight and Government Reform
Wednesday, March 7, 2018 at 2:00 p.m. 2154 Rayburn House Office Building
Game Changers: Artificial Intelligence Part II; Artificial Intelligence and the Federal Government
Chairman Hurd, Ranking Member Kelly, and members of the subcommittee, thank you for the opportunity to appear before you today. My name is Keith Nakasone, and I am the Deputy Assistant Commissioner for Acquisition Operations in the Office of Information Technology Category (ITC) at the U.S. General Services Administration (GSA). I have previously served in various roles at the Defense Information Systems Agency (DISA) and the Defense Information Technology Contracting Organization-Pacific (DITCO-PAC), so I have seen the growth of emerging technologies in government over the past twenty years.
Mr. Chairman, at the first hearing in this series, you stated that it was your hope agencies would use today’s discussion to inform Congress how we plan to “use AI [Artificial Intelligence] to spend taxpayer dollars wisely and make each individual's interactions with the government more efficient, effective, and secure.” I am pleased to be here today to tell you how GSA is doing just that, and how it plans to continue doing so.
GSA is proud of the innovative and cross-cutting work our team is undertaking to bring artificial intelligence to bear in the work we do, as well as the considerable effort we have engaged in to share best practices and use cases among partner agencies.
In its most recent strategic plan, the U. S. Government Accountability Office (GAO) identified AI as one of the “five emerging technologies [that] will potentially transform society.” What we have seen from our unique government-wide perch at GSA is that agencies are tremendously interested in AI and other emerging technologies - not because they are the latest fad, but because people recognize the potential to transform and simplify the way Americans interact with their government.
The modernization of the Federal government’s IT infrastructure and applications including emerging technologies such as Artificial Intelligence is an important priority for GSA. We are supporting government AI evaluation and adoption in four ways that I will introduce here and discuss further in my testimony.
- First, our Federal Acquisition Service (FAS) provides contracting vehicles and mechanisms including Schedule 70 and its associated programs, Startup Springboard and FAStlane, as well as several other government-wide acquisition contracts, which encourage competition and help connect agencies and businesses to allow government to efficiently procure the most effective new AI services.
- Second, we are piloting within our agency Robotic Process Automation (RPA) and other related technologies that are designed to augment our workforce to achieve more with less and establish a foundation for greater data-driven decision-making through AI.
- Third, through the interagency Emerging Citizen Technology Office (ECTO), we are helping support and coordinate government-wide development of citizen-facing AI programs, both public-facing as well as for internal agency use, with active participation from both the public and private sectors.
- Fourth, along with our private sector and federal agency partners, we are pursuing a greater understanding and alignment of IT modernization through cloud adoption, data services, and emerging technologies, including AI, that deliver the greatest benefit to the American people.
The Business of Artificial Intelligence Acquisitions in Government
Innovative companies across the country are using emerging technologies such as AI to build faster, smarter, and better products and services. GSA’s IT Schedule 70 contracts provide Federal, state, local, and tribal government agencies with access to over 7.5 million best-value IT and telecommunications products, services, and solutions from more than 4,600 pre-vetted vendors, including firms whose offerings use AI and similar technologies.
Since emerging technology businesses frequently tend to be startups or newer businesses, Schedule 70 offers two shortcuts - the FAStlane and Startup Springboard programs - as part of the Making It Easier (MIE) initiative, whose goal is to streamline the process for younger, innovative companies and suppliers to do business with government. This is a particular priority for GSA’s recently confirmed Administrator, Emily Murphy, who has a lengthy history of working with and helping small businesses, including as a staffer on the House Small Business Committee and as the Associate Administrator for Government Contracting at the U.S. Small Business Administration.
The FASt Lane program ensures Federal agencies have quicker access to emerging technologies and innovative AI suppliers. With FAStlane, suppliers get shorter processing times for IT Schedule 70 contract actions that directly support federal customer agency requirements, including 48-hour turnaround for contract modifications and turnaround in as quick as 45 days for new contract offers. Examples of AI are included in the Earth Observation SIN 132-41 with machine learning and algorithms that can be applied to globally scaled data, and the Cloud SIN 132-40 among others.
With Startup Springboard, AI vendors can now use their executives’ and key professionals' experience to substitute for two years of corporate experience. Startup Springboard has one primary objective: helping Federal agencies quickly gain access to the latest innovative technologies from fresh, vibrant private sector firms. Schedule 70 is the government’s largest IT contract vehicle, and provides a pathway for qualified innovative companies to partake in the $15 billion federal, state, and local IT contracting market.
- Beyond Schedule 70, a variety of emerging technology offerings, including those employing AI, are available under several other GSA acquisition vehicles:
- The Enterprise Infrastructure Solutions (EIS), a $50 billion information technology telecommunications vehicle and infrastructure solution, which is consolidating 93 separate contracts into a single contract with a 15-year period of performance.
- The Alliant/Alliant Small Business Governmentwide Acquisition Contracts (GWAC) provide flexible access to customized IT solutions for a large, diverse pool of industry partners.
- The VETS 2 (SDVOSB) Governmentwide Acquisition Contract (GWAC) was awarded Best-in-Class on October 26, 2017.
- The 8(a) STARS II Governmentwide Acquisition Contract (GWAC) is a small business set-aside that provides access to customized IT solutions from a large, diverse pool of 8(a) industry partners.
Robotic Process Automation and Data-Driven Decisionmaking
We are not only helping Federal agencies and businesses advance AI capabilities; our agency is bringing these best practices home to our own programs in order to provide the most effective and efficient services for our customers.
GSA’s Office of Government-wide Policy (OGP) has developed a new pilot using AI for Prediction of Regulatory Compliance, known as the Solicitation Review Tool (SRT). The SRT AI platform uses natural language processing, text mining, and machine learning algorithms to automatically predict whether federal solicitations posted on fbo.gov are compliant with Section 508 of the Rehabilitation Act and alert responsible parties of non-compliance so that corrective actions could be taken. Through independent review, the predictions have an accuracy rate of 95 percent. This innovation substantially alleviates the human resources needed to identify, audit, and enforce compliance.
The SRT platform is innovative because it helps GSA focus the limited resources available on the non-compliant solicitations identified and alert contracting staff to make the changes for compliance. The SRT tool is currently slated to go into production in cloud.gov in the spring of 2018. Future plans for the SRT AI platform include a scope expansion to predict whether solicitations contain other federal regulatory requirements such as cybersecurity or sustainability.
Furthermore, GSA launched two pilots in fiscal year 2018 exploring the use of Robotic Process Automation (RPA) and Distributed Ledger Technology (DLT), foundational technologies that can open our programs to better decision making through AI.
These pilots, in both our Office of the Chief Financial Officer (OCFO) and FAS, aim to:
- Increase GSA’s operational efficiency;
- Reduce cost;
- Improve processes;
- Increase accuracy; and
- Redeploy staff to higher-value functions.
GSA has established an inter-office RPA team, including OCFO, FAS, ECTO, and the Public Building Service, to provide governance of emerging technologies, including managing a common infrastructure and technical approach and helping business offices address process selection, prioritization, and implementation.
We look forward to continuing to share outcomes from these programs with our stakeholders and partners as they develop.
Emerging Citizen Technology Office Government-wide Services
GSA’s interagency ECTO unites more than 2,000 government managers from over 300 federal, state, and local agencies, and representatives from industry technology startups, small businesses, and leading research and civic organizations to develop government-wide IT modernization initiatives through the evaluation and strategic management of emerging technologies including AI, Robotic Process Automation, Blockchain, and Virtual and Augmented Reality. Our participants grow by at least six federal managers per day on average, underscoring the demand for more guidance and support for federal emerging technology programs.
Our AI program alone includes participation from 89 government agencies and growing, as well as a public-facing listserv to share updates and opportunities. This community includes experts from our military, civilian defense, and intelligence community. By sharing openly with private- sector innovators, startups, and new entrants in the field, U.S. businesses gain increased transparency into the modernization of federal information technology programs.
ECTO collaborates with experts on all aspects of appropriate use of AI including the Federal Privacy Council and the U.S. Data Cabinet, and serves on the U.S. National Science and Technology Council’s Subcommittee on Machine Learning and Artificial Intelligence.
ECTO also creates new opportunities for participation in our programs through mentorship in the Dcode emerging technology accelerator and U.S. Department of State’s Boldline public- private partnership accelerator.
We host an open source repository of potential use cases, programs, and resources for AI in the Federal government at Emerging.Digital.gov in the form of the Emerging Citizen Technology Atlas. This includes publicly reporting the action items from our monthly inter-agency meetings in order to help our partners in both the public and private sectors easily track the status of our AI initiatives, and soon will include reporting findings from our monthly “NewTech10” data call designed to provide an evolving snapshot of federal agencies’ programs and needs.
Public services are increasingly powered through the combination of greater access to data and practical advances in AI and Machine Learning that deliver solutions from research and development laboratories into the hands of programs everywhere.
From early experiences supporting an emergency response effort after a natural disaster that resulted in more requests received through digital services than could be processed by staff, ECTO finds the practical and immediate use cases for AI that help the American people today, from modernizing contact centers to empowering our teams with new and better ways to operationalize data for decision-making and operations.
For example, ECTO’s first initiative in AI for Citizen Services was to work with more than two dozen government agencies to use our open data to test new, automated 24-7 customer services in commercially available intelligent personal assistants.
Today, ECTO supports strategy and management of a wider spectrum of uses for AI based on proven need and demand from agencies and businesses, including workforce development, citizen customer service, and data analysis to target fraud, waste, and abuse.
We look forward to launching a new shared resource on Emerging.Digital.gov in the near future called “Emerging Technology Pathways to Acquisition.” This resource will increase transparency on ways agencies can acquire AI services by providing a simple roadmap businesses can use to open their solutions to government, including GSA schedules, prizes and competitions on Challenge.gov, and Joint Venture Partnerships.
ECTO also launched new initiatives this quarter to further support the advancement of AI in public services including an interagency Venture Capital Advisory Group, an Academic Research Outreach initiative to better open the breadth of government-wide programs to the leading U.S. research facilities, and an Education and Training pilot through GSA’s DigitalGov University that all agencies will be able to use.
Forecast: IT Modernization Alignment
The AI technology landscape includes interdependencies with many other technology areas. There are several foundational technology areas that must be aligned for government to successfully implement AI including data services, cybersecurity, and cloud. In this spirit, GSA is working to improve alignment of these technology areas internally, within GSA, and is providing improved guidance to partners across the Federal government.
This is in support of one of the key goals in GSA’s current Strategic Plan for Fiscal Years 2018- 2022: to improve the way Federal agencies buy, build, and use technology. Work is already underway in this area within and across several GSA programs.
First, through Data.gov and ECTO, we are learning how to improve the standardization and accessibility of the government open data that is a critical fuel of innovation. Through this work, we have solicited input from companies on how to improve data hosting so datasets are more easily digestible for AI and machine learning.
Second, through Federal Risk and Authorization Management Program (FedRAMP) and its standardized process, we help vendors utilizing AI achieve cyber security authorizations so that Federal agencies can easily and affordably buy them to improve citizen-facing services.
Third, through cloud.gov, Federal partners are able to easily acquire and deploy cloud-based AI services and products within their technology environment.
GSA will continue to develop strategies and approaches to further coordinate internal processes and programs for streamlining AI products and services to support the missions of Federal agencies. We see an opportunity for greater collaboration and coordination across government to help institutionalize the spread of these insights and best practices.
What are the challenges we face, as both an agency and a government? AI can and does mean different things to different people, different groups, and different agencies, in large part due to the plethora of very specific problems and solutions AI can be utilized for. While this is a challenge, it is also an opportunity, because it can help broaden our perspectives on AI, and give agencies a deeper understanding of the useful role this technology can play in so many spaces.
Additionally, when people hear “artificial intelligence,” their minds often wander to the realm of science fiction. Because it is difficult to explain what is and is not artificial intelligence, there could be a reluctance among some in society to embrace this technology. There is also a belief that AI is in the future, rather than in the present. Concerted effort by policymakers of all levels to help change this narrative will be critical in promoting acceptance and adoption of AI by more and more entities - which makes hearings like today’s all the more important.
Finally, while Americans now have access to more data than at any point in our history, the biggest challenge is how to harness this data make sense of these reams of information. GSA operates the Data.gov website, which is a centralized hub to access and use open Government datasets. At its launch in May of 2009, Data.gov had 47 datasets; it now features over 230,000 datasets. This provides citizens, businesses, researchers, and other agencies access to a vast array of critical information, and AI can help make sense of this data in a far quicker manner than is possible with human effort alone.
At end of day, GSA is essentially a shared service, and the Federal Acquisition Service is constantly seeking ways to make government faster, better, and smarter. AI is a tool that can expand the value proposition for federal agencies, vendors, and the American people alike.
Thank you again for the opportunity to be a part of this important hearing. I am confident that, together with you and the private sector, we can ensure that the United States of America retains the mantle of leadership on emerging technology for generations to come. I would be happy to answer any questions you might have.