Professional Documents
Culture Documents
b
AI ethics | A new imperative for businesses, boards, and C-suites
Introduction
For individuals, leaders, organizations, and society, Union’s General Data Protection Regulation
an increasingly urgent question is whether (GDPR) and the California Consumer Privacy
artificial intelligence (AI) can help us advance in Act have been promulgated, at least in part,
positive ways, or will it hinder us? to address these concerns.
01
AI ethics | A new imperative for businesses, boards, and C-suites
AI ethics, defined
02
AI ethics | A new imperative for businesses, boards, and C-suites
Considerations for
carrying out AI ethics
Conceptually, AI ethics applies to both the To understand where ethical issues could People, skills, organizational
goal of the AI solution, as well as each part arise and how in the future of work those models, and training.
of the AI solution. AI can be used to achieve issues might be avoided, it helps to organize Understand and monitor how
an unethical business outcome, even AI along four primary dimensions of concern the use of AI impacts experiences of both
though its parts— machine learning, deep (figure 1): employees and customers. Continuously
learning, NLP, and/or computer vision— assess how operating models, roles, and
were all designed to operate in an ethical Technology, data, and security. organizational models are evolving due
way. For example, an automated mortgage Look at the organization’s to the use of AI. Educate all levels of the
loan application system might include approach to the AI life cycle from workforce and implement training initiatives
computer vision and other tools designed to an ethical perspective, including the ways to retool or up-skill capabilities. Establish
read handwritten loan applications, analyze it builds and tests data and models into protocols to incentivize ethical behavior
the information provided by the applicant, AI-enabled solutions. Leadership in this and encourage ethical decisions along the
and make an underwriting decision based dimension comes from the organization’s AI life cycle. In this dimension, the human
on parameters programmed into the information, technology, data, security, and resources function shares responsibility with
solution. These technologies do not process privacy chiefs. learning and development teams, ethics
such data through an ethical lens—they just officers, and broader executive leadership.
process data. Yet if the mortgage company Risk management and
inadvertently programs the system with compliance. Public policy, legal, and
goals or parameters that discriminate Find out how the organization regulatory frameworks,
unfairly based on race, gender, or certain develops and enforces policies, procedures, and impact to society.
geographic information, the system could and standards for AI solutions. See how Finally, develop a sense of AI’s place in the
be used to make discriminatory loan they tie in with the organization’s mission, business environment. This includes the
approvals or denials. goals, and legal or regulatory requirements. level of acceptance AI has in government
The heads of risk, compliance, legal, and and culture. It also includes the direction
In contrast, an AI solution with an ethical ethics play a role in this dimension. that laws and regulations are taking with
purpose can include processes that lack regard to AI. Apply this information to the
integrity or accuracy toward this ethical effect AI is likely to have over time in terms
end. For example, a company may deploy of education, employment, income, culture,
an AI system with machine learning and other aspects of society.
capabilities to support the ethical goal of
nondiscriminatory personnel recruiting Te
ch The CEO, CRO, CCO, and CFO have
no
so or nd
processes. The company begins by using lo leadership roles across the first three
to ew l a
et ,
ci ks
gy
ct m ga
at
im ry cy
a,
criteria based on the best performers in the relies on leadership from politicians,
d ato oli
an
an gul c p
d
se
re bli
rit
performers may include biases based on y policymaking bodies.
past hiring characteristics (including FUTURE OF WORK
discriminatory criteria such as gender,
e
Pe m
nc
ia
pl
le de
m
, s ls
d
an
,o dt
t
en
an in
ag
at
an
io
na
sk
Ri
03
AI ethics | A new imperative for businesses, boards, and C-suites
Research and design The solution’s inherent risks (such as a computer A framework for what defines ethical use of AI and
vision application that captures and potentially data in the organization
misuses customers’ or employees’ images or
other personally identifiable information [PII]) are A cross-functional panel with representation from the
overlooked during the design phase business units as well as the technology, privacy, legal,
ethics, and risk groups
The solution’s intended purpose is not aligned to the
organization’s mission and values A data or AI ethics code of conduct by which
professionals must abide
Build and train The organization lacks appropriate ways to secure A process for determining where and how to obtain
consent from individuals whose data is used to train the data that trains the models
the AI model
Guidelines on where and how user consent becomes
A programmer builds bias (either consciously or a consideration in the training phase
unconsciously) into a model intended for an AI-
enabled solution, such as in the personnel recruiting Policies for where and how to build models and
example previously described whether to use open source technology
The data used to train an AI model has quality issues An assessment of ways that an AI solution can teach
itself behaviors out of sync with the organization’s
mission or values
Change and operate A chatbot (an AI application that can include cognitive A process for the organization to engage in
language capabilities) learns new behaviors that are continuous monitoring
inappropriate or offensive to customers
An assessment of ways that an AI-enabled solution
The organization is unable to quickly assess which can gain access to new forms of data
new data sources an AI-enabled solution has
recently accessed on its own A process for the business to update the board on
AI-related risks or issues
The organization lacks the ability to test and monitor
AI solutions Organizational thresholds or tolerance levels to help
determine whether to decommission an AI-enabled
solution
04
AI ethics | A new imperative for businesses, boards, and C-suites
05
AI ethics | A new imperative for businesses, boards, and C-suites
Addressing the AI
ethics imperative—
everyone’s
responsibility
With its machine learning, deep learning, life cycle. To be effective, each construct
NLP, and computer vision capabilities, AI should reflect an understanding of AI-
offers the exciting prospect of improving the related vulnerabilities along the previously
human condition. But there is a potential described four dimensions of risk.
dark side to AI that is hard to ignore.
It is too soon to tell where this journey will
The result is a new frontier in business lead businesses and their customers. But
ethics. Those involved with the the way forward is in sight. It begins with a
advancement of AI—including corporate top-level commitment to ethical leadership
boards, management teams, researchers, and a focus on technical and ethical literacy.
and engineers—face a growing imperative to With ethical guardrails to limit missteps
bring an ethical lens to what they design and along the way, everyone in an organization
build. This approach should be articulated can work together to produce solutions that
through organizational ethical constructs represent AI’s potential.
that apply throughout the AI product
06
AI ethics | A new imperative for businesses, boards, and C-suites
Contact us
Mary Galligan Courtney Parry
Managing Director Senior Manager
Deloitte Risk & Financial Advisory Deloitte Risk & Financial Advisory
Deloitte & Touche LLP Deloitte & Touche LLP
mgalligan@deloitte.com cparry@deloitte.com
Endnotes
1. “State of AI in the Enterprise, 2nd Edition,” Loucks, Davenport, and Schatsky, Deloitte Insights, Oct. 22,
2018, https://www2.deloitte.com/insights/us/en/focus/cognitive-technologies/state-of-ai-and-intelligent-
automation-in-business-survey.html.
2. https://www.merriam-webster.com/dictionary/artificial%20intelligence.
4. https://www.merriam-webster.com/dictionary/ethic.
07
This publication contains general information only and Deloitte is not, by means
of this publication, rendering accounting, business, financial, investment, legal,
tax, or other professional advice or services. This publication is not a substitute
for such professional advice or services, nor should it be used as a basis for any
decision or action that may affect your business. Before making any decision or
taking any action that may affect your business, you should consult a qualified
professional advisor.
Deloitte shall not be responsible for any loss sustained by any person who relies
on this publication.
About Deloitte
Deloitte refers to one or more of Deloitte Touche Tohmatsu Limited, a UK
private company limited by guarantee (“DTTL”), its network of member firms,
and their related entities. DTTL and each of its member firms are legally
separate and independent entities. DTTL (also referred to as “Deloitte Global”)
does not provide services to clients. In the United States, Deloitte refers to one
or more of the US member firms of DTTL, their related entities that operate
using the “Deloitte” name in the United States, and their respective affiliates.
Certain services may not be available to attest clients under the rules and
regulations of public accounting. Please see www.deloitte.com/about to learn
more about our global network of member firms.