On September 8, GSA held a workshop titled “Emerging Technology and Open Data for a More Accountable, Open Government” to solicit ideas from government, industry, and civil society leaders about how artificial intelligence, blockchain, and open data could make government more effective, more accountable, and more transparent. The goal of the workshop was to crowdsource ideas for the fourth U.S. National Action Plan for Open Government—the biennial commitments the United States makes to the Open Government Partnership, an initiative in which 75 national and 15 subnational governments have pledged to “promote transparency, empower citizens, fight corruption, and harness new technologies to strengthen governance.” While it remains to be seen just how the federal government will use these technologies, it is an encouraging sign that the executive branch is getting serious about using data-driven innovation to improve government.
Prior commitments in U.S. National Action Plans have rightfully emphasized the importance of open data for transparency, accountability, responsive government, but the forthcoming fourth National Action Plan will broaden this focus to leverage emerging data technologies as well. For example, none of the previous action plans include any commitments to use artificial intelligence or machine learning to improve open government. This is a timely change considering that many federal agencies have recently begun exploring how AI can improve their operations. In April 2017 for example, GSA launched a pilot project to help federal agencies make their data available for use with AI-powered personal assistants such as Amazon’s Alexa.
The workshop produced a wide variety of innovative proposals for the executive branch to consider. One such proposal was to encourage the adoption of accountable, open source AI software in the justice system for applications such as risk assessment and case management by publishing source code, training data, and machine learning tools and studying the impact of these tools in key metrics such as changes in incarceration rates and prison populations. Another proposal was for the federal government to spur the development of open standards for crisis mapping to improve humanitarian response efforts after a disaster. And presciently, another proposal called for the creation of a national blockchain-based personal identity management system to make it easier for people to interact with government services across multiple agencies. Such a system would likely have mitigated the damage caused by the recent Equifax hack, as over-reliance on social security numbers as a method for verifying an individual’s identity greatly increases the potential harm bad actors could inflict with stolen identifying information.
GSA should be commended for hosting this workshop as including a focus on emerging data technologies could help make the fourth National Action Plan considerably more impactful than previous plans. However, the federal government should also ensure it does not lose sight of the equally important commitments it made in past plans but has yet to complete. For example, the Treasury committed to requiring companies incorporated within the United States to disclose beneficial ownership data by pushing for legislation to require states to adopt this rule, yet Congress has yet to pass such legislation and as a result, domestic shell companies remain a safe haven for criminal activity. GSA’s workshop was nonetheless a beneficial exercise in promoting innovation in government, and the White House would do well to ensure that these and other innovative proposals make their way into the forthcoming National Action Plan.