With the internet becoming a more pronounced entity in
everyday life, invasion of privacy and security involving the unauthorised
collection, disclosure, or undisclosed use of personal information is becoming
a larger issue amongst consumers. Private information on consumers is now a
sought-after commodity which is routinely purchased, sold and traded between
groups. This information is typically gathered in two ways, either voluntarily
through purchases/website registration or involuntarily through browser cookies
which track online surfing behaviour. This information is then typically used
by businesses to tailor marketing strategies at specific groups of consumers
increasing efficiency and the effectiveness of their sites. The
confidentiality, integrity and privacy of personal data is a major issue in the
widespread adoption of new e-commerce technologies amongst consumers.
AI assisted e-commerce (commonly referred to as
voice-commerce) is the new and upcoming method of introducing e-commerce into
the relatively new AI assistants that exist on certain mobile/desktop operating
systems such as Windows, Android and IOS etc. and even exist in some dedicated
home technologies such as Google Home and Amazon Alexa.
This study focuses on aspects of introducing e-commerce
businesses into AI driven assistants as an infrastructure to communicate,
distribute and exchange information that lead to commercial transactions
between vendors and consumers. Additionally this study would like to identify
the main security and privacy concerns and the trustworthiness of AI assistant
driven e-commerce transactions and the effectiveness of the security methods
implemented within these AI assistants in ensuring confidentiality, integrity
and privacy in regards to e-commerce transactions.
This research intends to identify the direct factors
relating to consumers’ trust to adopt AI assistant driven e-commerce solutions
within the UK.
Therefore this study is set out to answer the following
Does consumer privacy and security concerns in
AI driven e-commerce transactions significantly affect adoption of AI driven
Does the trustworthiness/reliability of the
vendors supplying AI assistants affect consumers’ adoption of AI driven
What are the relationships between trust
beliefs, risk perceptions and security and privacy concerns, and how these
factors affect consumers’ intention to adopt AI driven e-commerce?
My knowledge in this field is limited so I will need to
collate information from other sources to ensure that this is a feasible study
in this field.
As defined by Hart (1998), literature reviews are important
to illustrating ways in which a topic has previously been investigated and can
possibly indicate where the topic of choice may be heading in future.
This literature review may help other authors explore the
effects of security and privacy concerns in technology on e-commerce.
This paper uses the recommendations of Webster & Watson
2002 in order to find relevant literature relating to this field. As privacy
and security concerns in relation to e-commerce is a widely discussed topic,
literature found in relation to the topic is widely available in multiple
contexts. Search terms within Google were used to find relevant reports
relating to this topic, using the search terms “security”, “privacy” and
“concerns” combined with “e-commerce”, “mobile”, “AI assistants”.
E-commerce has been gaining significant amounts of attention
over the years as businesses adopt fresh new ways of implementing e-commerce
solutions into everyday life, thus a lot of interesting studies revolving
around the topic of e-commerce have been developed.
The SAS which is a leader in analytics has determined that
62% of millennials would gladly use automated customer service systems if those
systems were to provide a better service, whilst 23% of customers would happily
let AI choose, purchase and deliver gifts to others. (4) This research is
specifically on the adoption of AI in businesses and not necessarily in the
adoption of AI assisted e-commerce solutions but the data is relevant to this
study as it gives us a basis for general adoption of AI technologies
particularly in the younger generations.
The review of Yazdanifard, Edres, Seyedi and Al-Huda 2011
concluded that security and privacy issues amongst consumers in regards to
e-commerce impacts growth and development within the e-commerce sector and emphasises
a need for increased security and privacy measures to be implemented and an
importance in ensuring healthy relations with online consumers in online
A review of Gurung, Raja 2015 found that privacy, security
concerns and trust had effects on perceptions of risk. The largest impact of
these being trust followed by privacy and security concerns, with behavioural
control and attitude having a positive effect on investment in e-commerce.
These topics talk mostly in relation to online e-commerce
and not in relation to AI assistants with ‘voice-commerce’, so this report fits
into the current discussions on the topic and can be used to expand the current
discussions relating to e-commerce and consumers’ trust adopting these
Research Design and Hypothesis
The objectives here are to identify the factors that
contribute to the willingness of consumers to engage in AI assisted e-commerce
transactions and study the relationships between these factors.
There will be three main categories of factors deemed to be
affecting consumers’ attitudes towards the AI assisted e-commerce system
The privacy of consumers’ personal data.
The reliability and trust of vendors supplying
AI driven e-commerce solutions.
H0: A consumers’ perception of AI driven e-commerce does not
affect adoption and use of these systems.
The following hypotheses are to be tested as part of this
H1: A consumers’ trust of AI driven e-commerce based online
transactions is positively related with the trustworthiness of the AI vendor
(Google, Amazon, etc).
H2: A consumers’ perceived lack of security in AI driven
e-commerce based transactions contributes negatively to their trust in the AI
H3: A consumers’ perceived lack of privacy in AI driven
e-commerce based transactions contributes negatively to their trust in the AI
investigation a questionnaire was developed that contained three different
The first section
consisted of questions designed to collect respondent’s personal information
including their gender and age.
The second section
of this questionnaire included questions relating to online purchasing habits
and adoption of AI driven e-commerce solutions. These questions are designed to
specifically collect frequency of internet use and frequency of online
purchases including through internet/AI driven solutions.
The third section
of the questionnaire consisted of questions designed to cover factors affecting
AI driven e-commerce security, trust and privacy. The questions in this section
employed the Likert scale ranging from 1 – Strongly Disagree to 5 – Strongly
questionnaires were used in the investigation, we ran a pilot test through a
series of informal interviews with a small selection of high level business
employees at IG to ensure that the questionnaires were properly operationalized
and properly implemented.
In the actual
implementation of this survey the target group for respondents was designed for
tech savvy developers at IG. 30 developers responded to the questionnaire with
60% of respondents being male and 40% female. The majority of the respondents
(95%) are between the ages of 20 and 30 with the remaining aged 31 and upwards.
From the findings within the survey 24 respondents strongly
agree that they often access the internet with 6 agreeing.
From this, respondents were asked about how many purchases
they make through the internet in an average month. The majority of respondents
stated that they purchased between 6-10 items each month through the internet.
With 3 respondents stating that they don’t purchase anything on average through
e-commerce solutions. These three were asked as to why they don’t purchase
through e-commerce sites, and they all stated that they felt that their
security and privacy concerns outweigh the benefits of e-commerce with one
responding that they prefer the idea of having the item immediately as part of
The questionnaire then asks whether users purchase items
through AI driven e-commerce solutions such as Amazon Alexa. This set of data
is drastically different from the online e-commerce data set,
We then asked the respondents who responded with no to
elaborate on their answers. We saw a large number of users not trusting the
newer AI assisted e-commerce technology over traditional e-commerce. The below
graph shows that of the 22 people a lot of peoples concerns with the AI
assisted implementation is revolving around security, trust and privacy
This is most likely relating to the technology being
relatively new and a distinct lack of widespread use amongst consumers.
Respondents who responded yes were asked after submission of
the questionnaire as to why they purchase items through these AI assisted
e-commerce solutions. Most of the responses revolved around the simplicity of
these systems, making it quicker, simpler and easier for users to purchase the
items they need/want without having to interact with traditional e-commerce
Respondents were then asked whether they felt that their
trust in an AI driven e-commerce vendor would positively affect their
purchasing decisions within that service. We found that many respondents stated
that their purchasing decisions with a system depend on their trust with the
vendor/provider of that service.
So, the data points towards vendors needing to build trust
within their userbase to ensure that their users feel comfortable using this
new implementation of the e-commerce system.
Next respondents were asked whether they felt their
purchasing decision would be negatively affected by AI driven e-commerce
vendors selling personal data. We found from the results that respondents would
be negatively affected by this, but there is a greater skew of people who feel
that this wouldn’t affect their decision to purchase through the service.
After the questionnaires were returned, we asked the
respondents who responded with ‘Neither Agree nor Disagree’ or ‘Disagree’ to
explain why they chose this option. Many of the respondents stated that they
don’t mind companies selling their personal data, with some users even stating
that they purposely use false information to counteract this kind of issue.
The respondents were then asked whether security issues
found within an AI driven e-commerce system would negatively affect their
purchasing decisions. The data that we got back clearly shows that this is the
most concerning of the three to consumers. With the data showing that all
respondents agreed that security issues would negatively affect their decisions
to use that service.
To conclude this report in regards to the original question,
I can see a definitive answer that shows that trust, privacy and security
concerns does affect consumers spending through AI assisted voice-commerce
Security issues were found to be the largest potential
factor affecting consumers from using AI assisted e-commerce systems. I can
definitively state that perceived security issues in systems negatively affect
consumers trust in the AI driven transaction.
Trust in AI driven e-commerce solutions were also found to
profoundly affect consumers purchasing decisions, with the majority of the
respondents agreeing/strongly agreeing that their trust in a system makes them
more likely to purchase through that service. 4 respondents neither agreed nor
disagreed that their trust in the vendor affected their purchasing decisions,
this could be down to a variety of factors including subjective views on e-commerce
systems, consumers only being able to purchase particular items through
particular vendors meaning that users have to use that vendor to purchase
Vendors selling personal data was found to have the widest skew
of data across the Likert scale, with ~76.7% of respondents agreeing that this
would negatively impact their purchasing decisions with an AI assisted
e-commerce vendor. ~13.4% of respondents neither agreed nor disagreed that this
affected their purchasing decision, with 10% of respondents disagreeing that
this would negatively affect their purchasing decision. After discussions with
the ~23.4% of respondents that didn’t agree, we found that many of these users
opted to using false information to protect themselves from these kinds of
These findings are important as they fit in line with
reports referenced in the literature review and solidifies the idea that
businesses should be putting consumers’ trust, security and privacy at the
forefront of their voice-commerce/e-commerce businesses. This data also fits
within the findings found in the Yazdanifard, Edres and Seyedi 2011 report
which describes the negative direct impact on the growth and development of
e-commerce based on trust, security and privacy as a lack of interest from the
consumer to engage in e-commerce would affect overall growth and development.
The data collected in this report in line with data
collected from resources used to research this topic, all point to a common
issue. Particularly in the Yazdanifard, Edres, Seyedi and Al-Huda report that
specifies that the biggest concerns to e-commerce development exist within consumers
lack of trust in the system, security concerns and privacy concerns, which
relates strongly to the data collected in this report.
These findings alongside other relevant findings should not
be taken lightly from a business perspective, as e-commerce growth and
development is directly affected by a customer’s trust in the system. This is
particularly apparent with the introduction of new ways for consumers to
purchase through e-commerce systems as new technologies and their success
depends greatly on consumers trusting these systems and using them.