IEEE Technology and Society Magazine - December 2021 - 54

Soft Law Governance of Artificial Intelligence
suggesting that these unknowns " could bring about
a whole new dimension in personal injury, property
damage and pure financial losses as well as third-party
liability risks, for instance, in product, environmental
and third-party liability " [22]. In acknowledging these
challenges, Munich Re went on to highlight that the
significant role insurance would have in facilitating the
positive impact of the nascent technology in parallel
with risk mitigation activities [22].
Swiss Re and Lloyds of London similarly produced
comprehensive reports exploring the evolving
state of the science and the role that the insurance
sector could have in managing uncertainties. In their
report, Swiss Re argued that nanomaterials should
not be considered de facto harmful and advocated
for the use of the precautionary principle [23]. The
state of the science had significantly evolved by the
time that Lloyds of London's published their reports
in 2007 and 2009, resulting in a more nuanced
approach to their discussion of benefits and risks.
The authors stressed, for example, the need to differentiate
between actual and perceived risk and
called for additional funding to help address the scientific
uncertainties [24]. Lloyds also expressed the
need to view nanotechnologies as being heterogeneous,
with different nanomaterials and applications
having varying levels of risk [24]. This sentiment is
important when looking at risk and AI.
Taken collectively, these reports largely focused
on the potential for mass tort litigation and the need
to avoid an analogous situation created by asbestos
fibers and their relationship to asbestosis and
mesothelioma [25]-[27]. The importance of robust
research was viewed as being crucial for informing
their underwriting activities and crafting of insurance
policies. For the most part, existing policies
and instruments were utilized, with some tweaks
where needed. However, some insurance companies
did innovate. For example, Lexington Insurance
Company, a Boston-based company owned by AIG,
rolled out their Lex NanoShield Policy in 2010. NanoShield
was designed specifically for entities whose
principal business was manufacturing nanomaterials
[28]. It appears that this type of coverage is no longer
offered, perhaps suggesting a lack of need. This may
be in part because, as of the time of this writing, there
had been no public reports of significant claims relating
to nano-related harms to human health.
While nanotechnologies and AI are associated
with different potential risks profiles, the former
54
illustrates that the early role insurance can play
when looking at risk linked to a rapidly evolving
and economically significant technology. When the
market leads in development, liability and insurance
quickly become the initial gatekeepers, especially in
the absence of command-and-control regulation.
Ai will disrupt, albeit to varying degrees, almost
every aspect of society-from how and where we
work, how we shop and move from place to place,
through to how we diagnose and treat disease. All
sectors will be affected as this rapidly evolving technology
continues to permeate society. The lack of
explicit regulatory tools should not imply that rules
and guidance do not exist to govern their use.
As demonstrated with nanotechnologies,
the
insurance sector can be one of the first " quasi-regulators "
with a powerful role to play in the management
of a technology. We argue that this is likely the case
with AI. Specifically, insurance companies (including
reinsurers) can wield notable power in shaping
and influencing the trajectory of how AI is developed
and brought into the market. Global companies such
as Lloyds, Swiss Re, and AIG-among others-will
be powerful actors in developing the AI landscape in
light of the ever-increasing body of claims data that
they hold relating to AI-based technologies. As new
claims are reported, these policies and premiums
will continue to evolve.
Inherently, liability and insurance policies balance
innovation and risk with consumer safety. Liability
serves as a sword, while insurance provides a shield.
Uniquely disruptive of this balance, AI forces market
participants to create new systems to cope with the
technology's risks. As such, we argue that insurance
schemes will play a pivotal role in the development of
these technologies going forward.

Acknowledgment
This work was supported in part by the Arizona
State University under a grant from the Charles Koch
Foundation.
 References
[1] B. Marr, " The key definitions of artificial intelligence
(AI) that explain its importance, " Forbes, Feb. 14,
2018. [Online]. Available: https://www.forbes. com/
sites/bernardmarr/2018/02/14/the-key-definitionsof-artificial-intelligence-ai-that-explain-itsimportance/#3859e0964f5d
IEEE
Technology and Society Magazine

https://www.forbes.com/sites/bernardmarr/2018/02/14/the-key-defiitions-of-articial-intelligence-ai-that-explain-its-importance/#3859e0964f5d https://www.forbes.com/sites/bernardmarr/2018/02/14/the-key-defiitions-of-articial-intelligence-ai-that-explain-its-importance/#3859e0964f5d https://www.forbes.com/sites/bernardmarr/2018/02/14/the-key-defiitions-of-articial-intelligence-ai-that-explain-its-importance/#3859e0964f5d https://www.forbes.com/sites/bernardmarr/2018/02/14/the-key-defiitions-of-articial-intelligence-ai-that-explain-its-importance/#3859e0964f5d

IEEE Technology and Society Magazine - December 2021

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