Artificial intelligence (AI) is already transforming how governments operate and it brims with both promise and peril. From law enforcement to social security claims adjudication to disease management and immigration, countless government agencies are already spending a considerable amount of money on AI for intelligent augmentation and delivery of government services, customer support, research, and regulatory analysis.
According to a recent report by Bloomberg Government, in less than two years since the White House published its executive order ‘Maintaining American Leadership in Artificial Intelligence,’ federal spending on AI is expected to cross $6 billion in 2021.
Designed right, automated decisioning systems in this new era of “algorithmic governance” can solve complex problems, create personalized experiences, reduce administrative burdens, and optimize resources for governments and private sector. However, given their ability to dynamically learn and evolve from data, black box AI algorithms are already leading to widespread discrimination, invasions of privacy, dangerous concentrations of power, diminished public trust, and the erosion of human autonomy.
Leveraging public sector spending to drive responsible AI adoption
At Responsible AI Institute (RAI) we are focused on leveraging public sector spending for an outsized impact on driving widespread adoption of Responsible AI that is trustworthy and ethical. By making adherence to responsible AI principles a prerequisite for bidding on public-sector AI contracts, we are helping government agencies:
- Ensure transparency, trustworthiness, and performance of their digital systems while building public trust and protecting privacy and civil liberties
- Set a responsible AI baseline for organizational AI acquisition practices and processes, operating structures, culture, guidelines, and policies
- Reimagine the Federal AI acquisition gateway and leverage it as a policymaking space for promoting trustworthy and ethical AI
- Create a fair and level playing field among AI technology and services providers to foster innovation in responsible AI
- Promote discussions surrounding AI regulations that balances visibility into software, algorithms, and data to mitigate potential public risks with the commercial need to protect intellectual property
- Establish operating models for responsible Procurement of AI for other countries to refine, adopt, and scale their adoption of Responsible AI
A four-year journey
Responsible AI Institute (RAI) was founded as a non-profit in 2017 to advance practical tools for advancing trustworthy and ethical AI. Early on, we recognized the immense potential of large public sector procurement of AI in driving awareness, shaping regulations, and promoting widespread adoption of Responsible AI.
Today’s collaboration announcement with the US Department of Defense Joint Artificial Intelligence Council (JAIC) on their pilot of a procurement review process, is a continuation of the journey in this area. The JAIC procurement effort through Tradewind will build upon the workshop and the continuing JAIC efforts as well as leverage the World Economic Forum and Responsible AI Institute partnership currently underway focused on developing a certification for the responsible use of AI.
RAI’s collaboration with the public sector began in 2018 with the Government of Canada on their Algorithmic Impact Assessment initiative to support the Canadian Treasury Board’s Directive on Automated Decision-Making. The goal of this initiative was to reduce risks from automated decisioning systems to Canadians and federal institutions, and drive more efficient, accurate, consistent, and interpretable decisions made pursuant to Canadian law.
Later in 2019, we collaborated with our partner The World Economic Forum on the AI Government Procurement Guidelines. Later that year we engaged in workshops with the University of Texas, Austin around AI Procurement Policy for State, Regional, and Local Governments. In 2020, this work led to a set of actionable procurement guidelines and guidebook for public-sector adoption of AI through government procurement called Procurement in a Box.
In spring 2021, we collaborated with the U.S. Government Accountability Office (GAO) and other members around the GAO developed an AI accountability framework. The GAO Accountability Framework for Federal Agencies and Other Entities is aimed at helping managers ensure accountability and responsible use of artificial intelligence in government programs and processes.
All this earlier work provided RAI Institute with a solid foundation to be chosen by the Department of Defense JAIC in a partnering role to help the US DoD innovate new procurement strategies and operating models for responsible Procurement of AI. We are honored to be selected and look forward to our contributions to help JAIC implement Responsible Artificial Intelligence in the Department of Defense in a manner that reflects the Department’s commitment to its ethical principles, including the protection of privacy and civil liberties.
The Road Ahead
The possibilities and perils of AI’s social disruptions have led the United States and many other countries around the world to promote and adopt the principles of trustworthy and responsible AI. There is growing recognition that the AI systems we design and deploy today are the systems we will live with tomorrow.
Senior Executives are now getting engaged in AI Procurement to establish Responsible AI guidelines and objectives, drive capacity building, and direct institutional buy-in. Many agencies and companies have not yet addressed this critical opportunity– and threat.
In the weeks and months ahead, our plan is to work with partners such as JAIC, IN3, EqualAI, World Economic Forum, Anthem, Schwarz Reisman Institute at the University of Toronto, and University of Texas at Austin and our community of thousands around the world to drive and guide conversations and provide tangible tools for Responsible AI Assessment and Responsible AI Certification.
In doing so, we will shape and accelerate strategic public and private sector AI procurement processes that make adherence to responsible AI principles a prerequisite for bidding on public and private sector AI contracts. We invite you to join our community to help us build a better world with Responsible AI.