Evolving Procurement for Artificial Intelligence Systems in Cities and Beyond

Array

As Artificial Intelligence Systems (AIS) technology is increasingly sourced from private companies for public use, existing procurement standards may fall short in proactively identifying harm or mitigating risks for citizens or users due to timing or budget constraints. Unfortunately, little research and interdisciplinary exchange exists to date on issues pertaining to AIS, public procurement, risk, transparency, and impact.

To develop meaningful interventions and strategies, it is essential to first generate a structured understanding of existing public procurement processes to identify how they can best intersect and support community needs. Cities are particularly important sites as they have become a uniquely important group of AIS clients.

As a way to address this need, IEEE Standards Association (IEEE SA) is collaborating on a project led by Dr. Mona Sloane of New York University and Dr. Rumman Chowdhury, CEO and Founder of Parity focused on evolving practices around procurement in AIS.  Beginning with key AIS needs of cities, the project will feature three roundtables with a public white paper that will be released in late Spring 2021.

“Responsible and ethical use of AI and machine learning technologies is of paramount importance as these technologies are more ubiquitous and influential in our daily lives,” notes Dr. Sloane. “The public sector provides a particularly sensitive and important application of data science, and we have to ensure that we can leverage public accountability to identify and mitigate harm – and procurement guidelines are a key mechanism by which this can be supported.”

The goal of the three procurement roundtables held with a group of experts from academia, business and policy is to:

  • identify concrete gaps in existing procurement and development processes
  • map them to the risks that will arise from widespread use of these technologies
  • create a new framework for public procurement guidelines and standards for public-use data science applications

By creating a novel interdisciplinary dialogue via these roundtables, this project seeds new research on the responsible use of data science in the public interest through procurement.

The future potential of this project is significant in that it will help create a specific and scalable research agenda beyond these initial roundtables and will lay the groundwork for a potential new IEEE standard on data science and public procurement through a white paper that will be produced at the end of the project.

“This work is unique and innovative because it is harnessing the essential benefits of multi-disciplinary collaboration which is a key focus of our work at IEEE,” notes Alpesh Shah, Senior Director of Global Business Strategy & Intelligence at IEEE SA. “Consensus is critical surrounding issues of procurement and AIS, especially for cities and municipalities looking to global experts to help provide uniform definitions, standards and certifications like our ECPAIS (Ethics Certification Program for Autonomous and Intelligent Systems, Ethically Aligned Design, and P7000 Standards work focused on ensuring responsible AIS design.”

“This project has the potential to be the first to set new public procurement standards that are based on a deep examination of definitions for key existing terms such as ‘transparency’, ‘accountability’, and ‘justice,’” noted Dr. Rumman Chowdhury of Parity. “We’re also excited to focus our initial research on how cities procure these technologies as cities are so immediately impactful on how AIS is adopted to directly influence a majority of the world’s population today.”

Particular focus will be placed on how these terms and associated due diligence through procurement can include marginalized populations. Lastly this work will be unique due to laying the groundwork for utilizing metrics for success of public interest technology procurement that go beyond traditional areas of risk mitigation and exponential growth, informed by the work of the IEEE 7010-2020 Standard.

Updates about this work along with the white paper generated from the workshops will be made available on an IEEE SA Website to be announced in the near future.

For more information on the project or to request a copy of the white paper, please contact Dr. Mona Sloane.

Share this Article