IEEE Technology and Society Magazine - December 2021 - 33

of review. They will plan for credibility and adoption
at the outset. Without any form of " constitutional "
legitimacy, AI soft law initiatives will need to earn
credibility through the fundamentals of participants,
process, and output. This means ongoing stakeholder
dialogue and effective demonstration of an output's
value to those stakeholders through adoption, evaluation,
and ongoing improvement. Current events and
the issues of credibility with tech, AI, and institutions
mean that work needs to be handled very thoughtfully
and with attention to specific parts of these fundamentals
around transparency, distributive fairness,
diffusion, and support of the wider ecosystem. This is
a delicate time for trust-based initiatives, particularly
those promoted to regulate a technology already perceived
to wield extraordinary power.
Finally, I should add one brief reality check.
Credibility is key to success for soft law initiatives.
But perfect credibility with all possible stakeholders
and purposes across so many different interests also
seems impossible. How much credibility is enough
for a given AI soft law initiative? The answer lies in
its definition-the work needs sufficient adoption
by implementers and the respect and support of
key stakeholders to play a trusted role in technology
governance. Answering the interests of a wide stakeholder
community is an important part of the soft
law balancing act to achieve credibility.
The evolution of an initiative often means that
stakeholders go through their own discovery of
how a project impacts their interests. Perspectives
change-initiatives themselves need to be responsive
to diverse stakeholder communities. No proponent
should go into an engagement with an
immovable view of what the output should be, even
as a strong vision of output needs to be part of the
early value and plan. Experienced participants know
that they will need to trade away important features
they care about, they will need to compromise, and
they still need to arrive at a credible outcome that
will be adopted and respected [71]. This too is part
of the ongoing dialogue across stakeholders.
SOft law iS an important tool in a wider landscape-proponents
should be cognizant of both
its power and limitations. Credibility is the lynchpin
connecting what is being developed with its ability to
garner respect from crucial stakeholders. No initiative
can claim credibility for itself-it must be earned. To
deliver on its promise, AI soft law will need to get a job
December 2021
done-it must direct and change behavior in ways
that meet societal goals. Doing so demands putting
credibility work squarely, strategically in the process,
from initial planning through adoption.

Acknowledgment
This work was supported in part by the Arizona
State University Artificial Intelligence Soft Law Workshop,
October 9, 2020.
I would like to thank the participants at the Arizona
State University Artificial Intelligence Soft Law
Workshop on October 9, 2020, for ideas and feedback,
and in particular Gary Marchant, Carlos Ignacio
Gutierrez, Katina Michael, Jason Matusow, Phil
Dawson, and Ashley Casovan for their encouragement
and suggestions.
 References
[1] U.S. House of Representatives Judiciary Committee,
" Online Platforms and Market Power, Part 6:
Examining the Dominance of Amazon, Apple,
Facebook, and Google, " Jul. 29, 2020.
[2] K.-F. Lee, AI Superpowers: China, Silicon Valley, and
the New World Order. Boston, MA, USA: Houghton
Mifflin Harcourt, 2018.
[3] European Commission, Robotics and Artificial
Intelligence Unit, " Artificial Intelligence, " Nov. 6, 2020.
Accessed: Nov. 27, 2020. [Online]. Available: https://
ec.europa.eu/digital-single-market/en/artificialintelligence
[4]
G. Marchant, " Soft law governance of artificial
intelligence, " AI Pulse, 2020. Accessed: Nov. 27,
2020. [Online]. Available: https://aipulse.org/softlawgovernance-of-artificial-intelligence/
[5]
R. Hagemann, J. Huddleston, and A. D. Thierer, " Soft
law for hard problems: The governance of emerging
technologies in an uncertain future, " Social Sci.
Res. Netw., Rochester, NY, USA, SSRN Scholarly
Paper 3118539, Feb. 2018. Accessed: Nov. 28,
2020. [Online]. Available: https://papers.ssrn.com/
abstract=3118539
[6] J. Villasenor, " Soft law as a complement to AI
regulation, " Brookings, Jul. 2020. Accessed: Nov. 28,
2020. [Online]. Available: https://www.brookings.edu/
[7] L. D. Brown, Creating Credibility: Legitimacy and
Accountability for Transnational Civil Society. Sterling,
VA, USA: Kumarian Press, 2008.
[8] P. Cairney, T. Heikkila, and M. Wood, Making Policy in
a Complex World. Cambridge, U.K.: Cambridge Univ.
Press, 2019.
33

https://ec.europa.eu/digital-single-market/en/artificial-intelligence https://ec.europa.eu/digital-single-market/en/artificial-intelligence https://ec.europa.eu/digital-single-market/en/artificial-intelligence https://www.aipulse.org/softlawgovernance-ofartificial-intelligence/ https://www.aipulse.org/softlawgovernance-ofartificial-intelligence/ https://papers.ssrn.com/abstract=3118539 https://papers.ssrn.com/abstract=3118539 https://www.brookings.edu/

IEEE Technology and Society Magazine - December 2021

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - December 2021