1 Which type of bias imposes a system's values on others?

2 A sales manager wants to improve their processes using

AI in Salesforce.
Which application of AI would be most beneficial?

3 What is the most likely impact that high- quality data

will have on customer relationships?

4 What are the key components of the data quality

standard?

5 What is the role of Salesforce's Trusted AI Principles in

the context of CRM systems?

6 What is a benefit of data quality and transparency as it

pertains to bias in generative AI?

7 A business analyst (BA) wants to improve business by

enhancing their sales processes and customer support.
Which AI applications should the BA use to meet their
needs?

8 How does AI within CRM help sales representatives

better understand previous customer interactions?
9 Why is it critical to consider privacy concerns when
dealing with AI and CRM data?

10 A data quality expert at Cloud Kicks wants to ensure that

each new contact contains at least an email address or
phone number.
Which feature should they use to accomplish this?

11 In the context of Salesforce's Trusted AI Principles, what

does the principle of Empowerment primarily aim to
achieve?

12 Cloud Kicks wants to use an AI model to predict the

demand for shoes using historical data on sales and
regional characteristics.
What is an essential data quality dimension to achieve
this goal?

13 A financial institution plans a campaign for preapproved

credit cards.
How should they implement Salesforce's Trusted AI
Principle of Transparency?

14 What is a key challenge of human-AI collaboration in

decision-making?
15 Which best describes the difference between predictive
AI and generative AI?

16 What is machine learning?

17 The Cloud Kicks technical team is assessing the

effectiveness of their AI development processes.
Which established Salesforce Ethical Maturity Model
should the team use to guide the development of
trusted AI solutions? The Cloud Kicks technical team is
assessing the effectiveness of their AI development
processes.
Which established Salesforce Ethical Maturity Model
should the team use to guide the development of
trusted AI solutions?

18 What is a potential source of bias in training data for AI

models?

19 Cloud Kicks wants to use AI to enhance its sales

processes and customer support.
Which capability should they use?

20 What can bias in AI algorithms in CRM lead to?

21 Which Einstein capability uses emails to create content

for Knowledge articles?
22 A system admin recognizes the need to put a data
management strategy in place.
What is a key component of a data management
strategy?

23 How does a data quality assessment impact business

outcomes for companies using AI?

24 What is an example of ethical debt?

25 A consultant conducts a series of Consequence Scanning

Workshops to support testing diverse datasets.
Which Salesforce Trusted AI Principle is being practiced?

26 Cloud Kicks wants to develop a solution to predict customers’

product interests based on historical data. The company
found that employees from one region use a text field to
capture the product category, while employees from all other
locations use a picklist.
Which dimension of data quality is affected in this
scenario?

27 A marketing manager wants to use AI to better engage

their customers.
Which functionality provides the best solution?

28 Which action should be taken to develop and implement

trusted generative AI with Salesforce's safety guideline
in mind?

29 Cloud Kicks learns of complaints from customers who

are receiving too many sales calls and emails.
Which data quality dimension should be assessed to
reduce these communication inefficiencies?
30 What is a potential outcome of using poor- quality data
in AI applications?

31 What role does data quality play in the ethical use of AI

applications?

32 Cloud Kicks wants to decrease the workload for its

customer care agents by implementing a chatbot on its
website that partially deflects incoming cases by
answering frequently asked questions.
Which field of AI is most suitable for this scenario?

33 Which type of bias results from data being labeled

according to stereotypes?

34 Salesforce defines bias as using a person's Immutable

traits to classify them or market to them.
Which potentially sensitive attribute is an example of an
immutable trait?

35 What are some key benefits of AI in improving customer

experiences in CRM?

36 How does an organization benefit from using AI to

personalize the shopping experience of online
customers?
37 Cloud Kicks is testing a new AI model. Which approach
aligns with Salesforce's Trusted AI Principle of
Incluslvity?

38 Cloud Kicks wants to develop a solution to predict

customers product interests based on historical data
The company found that employees from one region
use a text field to capture the product category, while
employees from all other locations use a plckllst.
Which data quality dimension is affected in this
scenario?

39 Cloud Kicks wants to implement AI features on its

salesforce Platform but has concerns about potential
ethical and privacy challenges.
What should they consider doing to minimize potential
AI bias?

40 Which features of Einstein enhance sales efficiency and

effectiveness?

41 What is the main focus of the Accountability principle in

Salesforce's Trusted AI Principles?

42 What is a sensitive variable that can lead to bias?

43 A Salesforce administrator creates a new field to capture

an order's destination country.
Which field type should they use to ensure data quality?
44 A customer using Einstein Prediction Builder is confused
about why a certain prediction was made. Following
Salesforce's Trusted AI Principle of Transparency, which
customer information should be accessible on the
Salesforce Platform

45 Which data does Salesforce automatically exclude from

marketing Cloud Einstein engagement model training to
mitigate bias and ethical concerns?

46 Cloud Kicks relies on data analysis to optimize its

product recommendation; however, CK encounters a
recurring Issue of Incomplete customer records, with
missing contact Information and incomplete purchase
histories. How will this incomplete data quality impact
the company’s operations?

47 A data quality expert at Cloud Kicks want to ensure that

each new contact contains at least an email address or
phone number. Which feature should they use to
accomplish this?

48 The Cloud technical team is assessing the effectiveness

of their AI development processes. Which established
Salesforce Ethical Maturity Model should the team use
to guide the development of trusted AI solution?
49 Cloud Kicks wants to use Einstein Prediction Builder to
determine a customer’s likelihood of buying specific
products; however, data quality is a …
How can data quality be assessed quality?

50 What can bias in AI algorithms in CRM lead to?

51 In the context of Salesforce’s Trusted AI Principles what

does the principle of Empowerment primarily aim to
achieve?
 A. Association C
 B. Automation
 C. Societal

 A. Lead scoring and opportunity forecasting A

 B. Data modeling and management
 C. Sales dashboards and reporting

 A. Improved customer trust and satisfaction A

 B. Increased brand loyalty
 C. Higher customer acquisition costs

 A. Accuracy, Completeness, Consistency A

 B. Reviewing, Updating, Archiving
 C. Naming, Formatting, Monitoring

 A. Outlining the technical specifications for AI integration C

 B. Providing a framework for AI data model accuracy
 C. Guiding ethical and responsible use of AI

 A. Chances of bias are aggravated. C

 B. Chances of bias are removed.
 C. Chances of bias are mitigated.

 A. Sales data cleansing and customer support data C

governance
 B. Machine learning models and chatbot predictions
 C. Lead scoring, opportunity forecasting, and case
classification

 A. Creates, localizes, and translates product descriptions B

 B. Provides call summaries
 C. Triggers personalized service replies
 A. Ensures compliance with laws and regulations A
 B. Confirms the data is accessible to all users
 C. Increases the volume of data collected

 A. Validation rule A
 B. Autofill
 C. Duplicate matching rule

 A. Empower users of all skill levels to build AI applications A

with clicks, not code.
 B. Empower users to solve challenging technical problems
using neural networks.
 C. Empower users to contribute to the growing body of
knowledge of leading AI research.

 A. Age B
 B. Reliability
 C. Volume

 A. Communicate how risk factors such as credit score can B

impact customer eligibility.
 B. Flag sensitive variables and their proxies to prevent
discriminatory lending practices.
 C. Incorporate customer feedback into the model’s
continuous training.

 A. Leads to more informed and balanced decision-making B

 B. Creates a reliance on AI, potentially leading to less
critical thinking and oversight
 C. Reduces the need for human involvement in decision-
making processes
A. Predictive AI and generative AI have the same capabilities but differ in C
the type of input they receive; predictive AI receives raw data whereas
generative AI receives natural language.
B. Predictive AI uses machine learning to classify or predict outputs from
its input data whereas generative AI does not use machine learning to
generate its output.
C. Predictive AI uses machine learning to classify or predict outputs from
its input data whereas generative AI uses machine leaning to generate
new and original output for a
given input.

 A. A data model used in Salesforce B

 B. AI that can grow its intelligence
 C. AI that creates new content

 A. Ethical AI Prediction Maturity Model B

 B. Ethical AI Practice Maturity Model
 C. Ethical AI Process Maturity Model

 A. The data is collected in real time from source systems. C

 B. The data is collected from a diverse range of sources
and demographics.
 C. The data is skewed toward a particular demographic or
source.

 A. Sales Path and Automated Case Escalations B

 B. Einstein Lead Scoring and Case Classification
 C. Dashboard of Current Leads and Cases

 A. Ethical challenges in CRM systems A

 B. Advertising cost increases
 C. Personalization and targeted marketing changes

 A. Predict C
 B. Discover
 C. Generate
 A. Naming Convention C
 B. Color Coding
 C. Data Backup

 A. Provides a benchmark for AI predictions A

 B. Accelerates the delivery of new AI solutions
 C. Improves the speed of AI recommendations

 A. Violating a data privacy law and failing to pay fines C

 B. Delaying an AI product launch to retrain an AI data
model
 C. Launching an AI feature after discovering a harmful bias

 A. Accountability B
 B. Inclusivity
 C. Transparency

 A. Accuracy C
 B. Completeness
 C. Consistency

 A. Bring Your Own Model C

 B. Journey Optimization
 C. Einstein Engagement

 A. Be transparent when AI has created and autonomously C

delivered content.
 B. Develop right-sized models to reduce our carbon
footprint.
 C. Create guardrails that mitigate toxicity and protect PII.

 A. Duplication A
 B. Consent
 C. Usage
  A. AI models may produce biased or erroneous results. A
 B. AI models become more interpretable.
 C. AI model training becomes slower and less efficient.

A. High-quality data is essential for ensuring unbiased and fair AI decisions, A

promoting ethical use, and preventing discrimination.
B. Low-quality data reduces the risk of unintended bias as the data is not overfitted
to demographic groups.
C. High-quality data ensures the presence of demographic attributes required for
personalized campaigns.

 A. Natural language processing A

 B. Predictive analytics
 C. Computer vision

 A. Association B
 B. Societal
 C. Interaction

 A. Financial status A
 B. Nickname
 C. Email address

 A. Improves CRM security protocols, safeguarding sensitive B

customer data from potential breaches and threats
 B. Streamlines case management by categorizing and
tracking customer support cases, identifying topics, and
summarizing case resolutions
 C. Fully automates the customer service experience,
ensuring seamless automated interactions with customers

 A. Customers are more likely to share personal B

information with a site that personalizes their experience.
 B. B. Customers are more likely to be satisfied with their
shopping experience.
 C. Customers are more likely to visit competitor sites that
personalize their experience.
A. Test only with data from a specific region or demographic
to limit the risk of data leaks.
B. B. Rely on a development team with uniform backgrounds
to assess the potential societal implications of the model.
C. C. Test with diverse and representative datasets
appropriate for how the model will be used.

A. Completeness B
B. Accuracy
C. consistency

A. Integrate AI models that auto correct biased data B

B. Implement Salesforce’s Trusted AI principles
C. Use demographic data to identify minority groups

A. Opportunity List View, Lead List View, Account List view C

B. B. Opportunity Scoring, Opportunity List View,
Opportunity Dashboard
C. C. Opportunity Scoring, Lead Scoring, Account Insights

A. Safeguarding fundamental human rights and protecting B

sensitive data
B. B. Taking responsibility for one's actions toward
customers, partners, and society
C. C. Ensuring transparency In Al-driven recommendations
and predictions

A. Education level C
B. County
C. Gender

A. Text B
B. Picklist
C. number
A. An explanation of how Prediction Builder works and a link to B
Salesforce's Trusted AI Principles
B. An explanation of how Prediction Builder works and a link to
Salesforce's Trusted AI Principles
C. A marketing article of the product that clearly
outlines the oroduct's capabilities and features

A. Geographic B
B. Geographic
C. Cryptographic

A. The accuracy of product recommendations is hindered. A

B. The diversity of product recommendations Is Improved.
C. The response time for product recommendations is stalled.

A. Autofill C
B. Duplicate matching rule
C. Validation rule

A. Ethical AI Prediction Maturity Model B

B. Ethical AI Process Maturity Model
C. Ethical AI practice Maturity Model
A. Build a Data Management Strategy. C
B. Build reports to expire the data quality.
C. Leverage data quality apps from AppExchange

A. Personalization and target marketing changes C

B. Advertising cost increases
C. Ethical challenges in CRM systems

A. Empower users to off all skill level to build AI application with A

clicks, not code.
B. Empower users to contribute to the growing body of knowledge
of leading AI research.
C. Empower users to solve challenging technical problems using
neural networks.
?????????
"A data model is a machine learning feature used in Salesforce. A data model is a representation
or abstraction of a real-world phenomenon or process using data structures and algorithms. A
data model can be used to describe, analyze, or predict various aspects of the phenomenon or
process using machine learning techniques."

"Einstein Generate uses emails to create content for Knowledge articles. Einstein Generate is a
natural language generation (NLG) feature that can automatically write summaries,
descriptions, or recommendations based on data or text inputs. For example, Einstein Generate
can analyze email conversations between agents and customers and generate draft articles for
the Knowledge base."
"Einstein Engagement provides the best solution for a marketing manager who wants to use AI
to better engage their customers. Einstein Engagement is a feature that uses AI to optimize
email marketing campaigns by providing insights and recommendations on the best time,
frequency, content, and subject lines to send emails to each customer. Einstein Engagement
can help increase customer engagement, retention, and loyalty by delivering
personalized and relevant messages.

Dans le contexte des directives de sécurité de Salesforce, la création de garde- fous pour la
toxicité et la protection des PII est une considération plus directe et spécifique.
L'option B, développer des modèles de la bonne taille pour réduire l'empreinte carbone, est une
considération valable pour la durabilité environnementale

"Duplication is the data quality dimension that should be assessed to reduce communication
inefficiencies. Duplication means that the data contains multiple copies or instances of the same
record or value. Duplication can cause confusion, errors, or waste in data analysis and
processing. For example, duplication can lead to communication inefficiencies if customers
receive multiple calls or emails from different sources for the same purpose."
"Societal bias results from data being labeled according to stereotypes. Societal bias is a type of
bias that reflects the assumptions, norms, or values of a specific society or culture. For example,
societal bias can occur when data is labeled based on gender, race, ethnicity, or religion
stereotypes."

"Financial status is an example of an immutable trait. Immutable traits are characteristics that
are inherent, fixed, or unchangeable. For example, financial status is an immutable trait because
it is determined by factors beyond one's control, such as birth, inheritance, or economic
conditions. Nickname and email address are not immutable traits because they can be changed
by choice or preference."

"Streamlining case management by categorizing and tracking customer support cases,

identifying topics, and summarizing case resolutions are some key benefits of AI in improving
customer experiences in CRM. AI can help automate and optimize various aspects of customer
service, such as routing cases to the right agents, providing relevant information or suggestions,
and generating reports or insights. AI can also help enhance customer satisfaction and loyalty by
reducing wait times, improving response quality, and providing personalized solutions."

"An organization benefits from using AI to personalize the shopping experience of online
customers by increasing customer satisfaction. AI can help provide customized and relevant
product recommendations, offers, or content based on the customers' preferences, behavior,
or needs. AI can also help create a more engaging and interactive shopping experience by using
natural language processing (NLP) or computer vision techniques. Personalized shopping
experiences can improve customer satisfaction by meeting their expectations, needs, and
interests."
"Testing with diverse and representative datasets appropriate for how the model will be used
aligns with Salesforce's Trusted AI Principle of Inclusivity. Inclusivity means that AI systems
should be designed and developed with respect for diversity and inclusion of different
perspectives, backgrounds, and experiences. Testing with diverse and representative datasets
can help ensure that the models are fair, unbiased, and representative of the target population
or domain."

Explanation:
"Consistency is the data quality dimension that is affected in this scenario. Consistency means
that the data values are uniform and follow a common standard or format across different
records, fields, or sources. Inconsistent data can cause confusion, errors, or duplication in data
analysis and processing. For example, using different field types for the same attribute can
affect the consistency of the data."

"Implementing Salesforce's Trusted AI Principles is what Cloud Kicks should consider doing to
minimize potential AI bias. Salesforce's Trusted AI Principles are a set of guidelines and best
practices for developing and using AI systems in a responsible and ethical way. The principles
include Accountability, Fairness & Equality, Transparency & Explainability, Privacy & Security,
Reliability & Safety, Inclusivity & Diversity, Empowerment & Education."

"Opportunity Scoring, Lead Scoring, Account Insights are features of Einstein that enhance sales
efficiency and effectiveness. Opportunity Scoring and Lead Scoring use predictive models to
assign scores to opportunities and leads based on their likelihood to close or convert. Account
Insights use natural language processing (NLP) to provide relevant news and insights about
accounts based on their industry, location, or events."

Explanation:
"The main focus of the Accountability principle in Salesforce's Trusted AI Principles is taking
responsibility for one's actions toward customers, partners, and society. Accountability means
that AI systems should be designed and developed with respect for the impact and
consequences of their actions on others. Accountability also means that AI developers and users
should be aware of and adhere to the ethical, legal, and regulatory standards and expectations
of their industry and domain."

"Gender is a sensitive variable that can lead to bias. A sensitive variable is a variable that can
potentially cause discrimination or unfair treatment based on a person's identity or
characteristics.
For example, gender is a sensitive variable because it can affect how people are perceived,
treated, or represented by AI systems."

"A picklist field type should be used to ensure data quality for capturing an order's destination
country. A picklist field type allows the user to select one or more predefined values from a list.
picklist field type can ensure data quality by enforcing consistency, accuracy, and completeness
of the data values."
"An explanation of the prediction's rationale and a model card that describes how the model
was created should be accessible on the Salesforce Platform following Salesforce's Trusted AI
Principle of Transparency. Transparency means that AI systems should be designed and
developed with respect for clarity and openness in how they work and why they make certain
decisions. Transparency also means that AI users should be able to access relevant information
and documentation about the AI systems they interact with."

Demographic data is the data that Salesforce automatically excludes from Marketing Cloud
Einstein engagement model training to mitigate bias and ethical concerns. Demographic data is
data that describes the characteristics of a population or a group of people, such as age, gender,
race, ethnicity, income, education, or occupation. Demographic data can lead to bias if it is used
to discriminate or treat people differently based on their identity or attributes.
Demographic data can also reflect existing biases or stereotypes in society or

culture, which can affect the fairness and ethics of AI systems. Salesforce excludes demographic
data from Marketing Cloud Einstein engagement model training to mitigate bias and ethical
concerns by ensuring that the models are
based on behavioral data rather than personal data.
The correct answer is A. The accuracy of product recommendations is hindered. Incomplete
customer records, with missing contact information and incomplete purchase histories, can
affect the quality of data analysis and lead to poor product recommendations. Data analysis
relies on having complete and accurate data to identify patterns, trends, and preferences.
Without sufficient data, the product recommendations may not match the customer’s needs,
interests, or expectations, resulting in lower customer satisfaction and retention.
The incomplete data quality will impact the company’s operations by hindering the accuracy of
product recommendations. Incomplete data means that the data is missing some values or
attributes that are relevant for the AI task.
Incomplete data can affect the performance and reliability of AI models, as they may not have
enough information to learn from or make accurate predictions. For example, incomplete
customer records can affect the quality of product recommendations, as the AI model may not
be able to capture the customers’ preferences, behavior, or needs.

The correct answer is C. Validation rule. A validation rule is a feature that allows you to define
criteria for data entry and prevent users from saving records that do not meet the criteria. In
this case, a validation rule can be used to ensure that each new contact contains at least an
email address or phone number by checking if either of these fields is populated. If not, the
validation rule will display an error message and prevent the user from saving the record.
A validation rule should be used to ensure that each new contact contains at least an email
address or phone number. A validation rule is a feature that checks the data entered by users
for errors before saving it to Salesforce. A validation rule can help ensure data quality by
enforcing certain criteria or
conditions for the data values.

The Ethical AI Process Maturity Model is the established Salesforce Ethical Maturity Model that
the Cloud technical team should use to guide the development of trusted AI solutions. The
Ethical AI Process Maturity Model is a framework that helps assess and improve the ethical and
responsible practices and processes involved in developing and deploying AI systems. The
Ethical AI Process Maturity Model consists of five levels of maturity: Ad Hoc, Aware, Defined,
Managed, and Optimized. The Ethical AI Process Maturity Model can help guide the
development of trusted AI solutions by providing a roadmap and
best practices for achieving higher levels of ethical maturity.
Leveraging data quality apps from AppExchange is how data quality can be assessed. Data
quality is the degree to which data is accurate, complete, consistent, relevant, and timely for
the AI task. Data quality can affect the performance and reliability of AI systems, as they depend
on the quality of the data they use to learn from and make predictions. Leveraging data quality
apps from AppExchange means using third-party applications or solutions that can help
measure, monitor, or improve data quality in Salesforce.

The correct answer is C. Ethical challenges in CRM systems. Bias in AI algorithms in CRM can
lead to ethical challenges in CRM systems, such as unfair or discriminatory treatment of
customers, violation of customer privacy, or erosion of customer trust. Bias in AI algorithms can
result from various factors, such as biased data, biased models, or biased outcomes. Bias can
affect the quality and accuracy of the AI predictions and recommendations, and ultimately
impact the customer experience and satisfaction. Therefore, it is important to identify and
mitigate bias in AI algorithms in CRM and ensure that they are aligned with ethical principles
and values.
Bias in AI algorithms in CRM can lead to ethical challenges in CRM systems. Bias means that AI
algorithms favor or discriminate certain groups or outcomes based on irrelevant or unfair
criteria. Bias can affect the fairness and ethics of CRM systems, as they may affect how
customers are perceived, treated, or represented by AI algorithms. For example, bias can lead
to ethical challenges in CRM systems if AI algorithms make inaccurate or harmful predictions or
recommendations based on customers’ identity or characteristics.

The correct answer is A. Empower users of all skill levels to build AI applications with clicks, not
code. The principle of Empowerment is one of the four Trusted AI Principles that Salesforce
follows to ensure that its AI solutions are ethical, transparent, accountable, and fair. The
principle of Empowerment aims to achieve the following goals:
 Enable users of all skill levels to build AI applications with clicks, not code, using declarative
tools and pre-trained models.
 Provide users with guidance and best practices on how to use AI responsibly and
effectively.
52 How does the "right of least privilege" reduce the risk of
handling sensitive personal data?

53 What is the best method to safeguard customer data

privacy?

54 What is the key difference between generative and

predictive AI?

55 Cloud Kicks implements a new product recommendation

feature for its shoppers that recommends shoes of a
given color to display to customers based on the color of
the products from their purchase history. Which type of
bias is most likely to be encountered in this scenario?
56 What is the role of Salesforce Trust AI principles in the
context of CRM system?
 A. By limiting how many people have access to data A
B. B. By reducing how many attributes are collected
C. C. By applying data retention policies

A. Automatically anonymize all customer data. B

B. B. Track customer data consent preferences.
C. C. Archive customer data on a recurring schedule.

A.Generative AI creates new content based on existing data and

predictive AI analyzes existing data.
B. Generative AI finds content similar to existing data and
predictive AI analyzes existing data.
C. Generative AI analyzes existing data and predictive AI creates
new content based on existing data.

A. Confirmation A
B. Survivorship
D. Societal
A. Guiding ethical and responsible use of AI A
B. Providing a framework for AI data model accuracy
C. Outlining the technical specifications for AI integration
 Educate users on the benefits and limitations of AI and how to interpret and act on the AI
outputs.
The principle of Empowerment primarily aims to achieve empowering users of all skill levels to
build AI applications with clicks, not code. Empowerment is one of the Trusted AI Principles that
states that AI systems should be designed and developed with respect for the empowerment
and education of humans. Empowering users means enabling users to access, use, and benefit
from AI systems regardless of their technical expertise or background. For example,
empowering users means providing tools and platforms that allow users to build AI applications
with clicks, not code, such as Einstein Prediction Builder or
Einstein Discovery.

"The "right of least privilege" reduces the risk of handling sensitive personal data by limiting
how many people have access to data. The "right of least privilege" is a security principle that
states that each user or system should have the minimum level of access or privilege necessary
to perform their tasks or functions. The "right of least privilege" can help protect sensitive
personal data from unauthorized access, misuse, or leakage."

"Tracking customer data consent preferences is the best method to safeguard customer data
privacy. Data privacy is the right of individuals to control how their personal data is collected,
used, shared, or stored by others. Tracking customer data consent preferences means
respecting and honoring the choices and preferences of customers regarding their personal
data. Tracking customer data consent preferences can help ensure compliance with data
privacy laws
and regulations, as well as build trust and loyalty with customers."

"The key difference between generative and predictive AI is that generative AI creates new
content based on existing data and predictive AI analyzes existing data. Generative AI is a type
of AI that can generate novel content such as images, text, music, or video based on existing
data or inputs. Predictive AI is a type of AI that can analyze existing data or inputs and make
predictions or recommendations based on patterns or trends."

The correct answer is A. Confirmation bias. Confirmation bias is a type of bias that occurs when
people tend to seek, interpret, and remember information that confirms their existing beliefs or
preferences. In this scenario, confirmation bias can affect the product recommendation feature
by only recommending shoes of a given color to customers based on the color of the products
from their purchase history. This can limit the diversity and variety of the recommendations and
prevent customers from discovering other products that they might like or need. Confirmation
bias can also reduce customer satisfaction and loyalty if the recommendations are not relevant
or
personalized

Confirmation bias is most likely to be encountered in this scenario. Confirmation bias is a type of
bias that occurs when data or information confirms or supports one’s existing beliefs or
expectations. For example, confirmation bias can occur when a product recommendation
feature only recommends shoes of a given color based on the customer’s purchase history,
without considering other factors or preferences that may influence their
choice.
The correct answer is A. Guiding ethical and responsible use of AI. Salesforce Trust AI principles
are a set of principles that Salesforce follows to ensure that its AI solutions are ethical,
transparent, accountable, and fair. The principles are:
 Empowerment: Enable users of all skill levels to build AI applications with clicks, not code,
and provide guidance and best practices on how to use AI responsibly and effectively.
 Transparency: Provide users with clear and understandable information about how the AI
works, what data it uses, and what outcomes it produces.
 Accountability: Ensure that the AI is reliable, secure, and compliant with laws and
regulations, and that users can provide feedback and report issues.
 Fairness: Prevent and mitigate bias in the AI and ensure that the AI respects the diversity
and dignity of the customers.
The role of Salesforce Trust AI principles in the context of CRM system is to guide the ethical
and responsible use of AI, and to ensure that the AI solutions enhance the customer experience
and satisfaction, while respecting the customer privacy and trust.
“The role of Salesforce Trust AI principles in the context of CRM systems is guiding ethical and
responsible use of AI. Salesforce Trust AI principles are a set of guidelines and best practices for
developing and using AI systems in a responsible and ethical way. The principles include
Accountability, Fairness & Equality, Transparency & Explainability, Privacy & Security, Reliability
& Safety, Inclusivity & Diversity, Empowerment & Education. The principles aim to ensure that
AI systems are aligned with the values and interests of customers, partners, and society.”
57 A developer is tasked with selecting a suitable dataset
for training an AI model in Salesforce to accurately
predict current customer behavior. What Is a crucial
factor that the developer should consider during
selection?

58 Which Einstein capacity uses emails to create content

for Knowledge articles?

59 What should organizations do to ensure data quality for

their AI initiatives?

60 Cloud Kicks wants to create a custom service analytics

application to analyze cases in Salesforce. The
application should rely on accurate data to ensure
efficient case
resolution. Which data quality dimension Is essential for
this custom application?

61 What is an example of Salesforce’s Trusted AI Principle

of Inclusivity in practice?
62 How is natural language processing (NLP) used in the
context of AI capabilities?

63 Cloud Kicks discovered multiple variations of state and

country values in contact records. Which data quality
dimension is affected by this issue?

64 What are some of the ethical challenges associated with

AI development?
65 A healthcare company implements an algorithm to
analyze patient data and assist in medical diagnosis.
Which primary role does data Quality play In this AI
application?

66 Cloud Kicks wants to optimize its business operations by

incorporating AI into its CRM. What should the company
do first to prepare its data for use with AI?
67 What are the three commonly used examples of AI in
CRM?

68 What is a benefit of a diverse, balanced, and large

dataset?

69 What is a key benefit of effective interaction between

humans and AI systems?

70 What is a key characteristic of machine learning in the

context of AI capabilities?

71 Cloud Kicks wants to ensure that multiple records for the same
customer are removed in Salesforce.
Which feature should be used to accomplish this?
72 An administrator at Cloud Kicks wants to ensure that a
field is set up on the customer record so their preferred
name can be captured.
Which Salesforce field type should the administrator use
to accomplish this?

73 What is a possible outcome of poor data quality?

74 To avoid introducing unintended bias to an AI model,

which type of data should be omitted?

75 What is an implication of user consent in regard to AI

data privacy?

76 How does data quality impact the trustworthiness of Al-

driven decisions?
A. Number of variables ipn the dataset B
B. Size of the dataset
C. Age of the dataset

A. Generate A
B. Discover
C. Predict

A. Collect and curate high-quality data from reliable sources. A

B. Rely on AI algorithms to automatically handle data quality issues.
C. Prioritize model fine-tuning over data quality improvements.

A. Consistency
B. Duplication
C. Age

A. Testing models with diverse datasets A

B. Striving for model explain ability
C. Working with human rights experts
A. To cleanse and prepare data for AI implementations C
B. To interpret and understand programming language
C. To understand and generate human language

A. Usage C
B. Accuracy
C. Consistency

A. Potential for human bias in machine learning algorithms and the A

lack of transparency in AI decision- making processes
B. Implicit transparency of AI systems, which makes It easy for
users to understand and trust their decisions
C. Inherent neutrality of AI systems, which eliminates any potential
for human bias in decision-making
A. Enhanced accuracy and reliability of medical predictions and A
diagnoses
B. Ensured compatibility of AI algorithms with the system’s
Infrastructure
C. Reduced need for healthcare expertise in interpreting AI outouts

A. Remove biased data.

B. Determine data availability.
C. Determine data outcomes.
A. Predictive scoring, reporting, Image classification
B. Predictive scoring, forecasting, recommendations
C. Einstein Bots, face recognition, recommendations

A. Training time
B. Data privacy
C. Model accuracy

A. Leads to more informed and balanced decision making A

B. B. Alerts humans to the presence of biased data
C. C. Reduces the need for human involvement

A. Uses algorithms to learn from data and make decisions A

B. B. Relies on preprogrammed rules to make decisions
C. C. Can perfectly mimic human intelligence and decision-
making

A. Duplicate management A
B. trigger deleting of old records
C. standardized field names
A. Multi select picklist B
B. text
C. rich text area

A. AI models maintain accuracy but have slower response B

times.
B. B. Biases in data can be inadvertently learned and
amplified by AI systems.
C. C. AI predictions become more focused and less
robust.

A. Transactional C
B. engagement
C. demographic

A. AI ensures complete data privacy by automatically B

obtaining user consent.
B. B. AI infringes on privacy when user consent is not
obtained.
C. C. AI operates Independently of user privacy and consent.

A. The use of both low-quality and high-quality data can improve B

the accuracy and reliability of AI- driven decisions.
B. B. High-quality data improves the reliability and credibility of Al-
driven decisions, fostering trust among users.
C. C. Low-quality data reduces the risk of overfitting the model,
improving the trustworthiness of the
predictions.
The size of the dataset is a crucial factor that the developer should consider during selection.
The size of the dataset refers to the amount or volume of data available for training an AI
model. The size of the dataset can affect the feasibility and quality of the AI model, as well as
the choice of AI techniques and tools. The size of the dataset should be large enough to provide
sufficient information for the AI model to learn from and generalize well to new data.

The correct answer is A. Generate. Einstein Generate is a capacity that uses emails to create
content for Knowledge articles. Einstein Generate can analyze the emails from customers and
agents and extract the relevant information and insights. It can then use natural language
generation to create draft articles that can be reviewed and published by the agents. Einstein
Generate can help to automate the process of creating Knowledge articles and improve the
quality and consistency of the content.
Einstein Generate uses emails to create content for Knowledge articles. Einstein Generate is a
natural language generation (NLG) feature that can automatically write summaries,
descriptions, or recommendations based on data or text inputs. For example, Einstein Generate
can analyze email conversations between agents and customers and generate draft articles for
the Knowledge base.

The correct answer is A. Collect and curate high-quality data from reliable sources. Data quality
is a key factor for the success of AI initiatives, as it affects the performance and accuracy of the
AI models and solutions. Data quality refers to the characteristics of the data, such as
completeness, correctness, consistency, timeliness, and relevance. To ensure data quality for
their AI initiatives, organizations should collect and curate high-quality data from reliable
sources, such as internal databases, external partners, or public datasets. They should also
implement data quality processes and tools, such as data cleansing, data validation, data
integration, and data monitoring, to identify and resolve any data quality issues and maintain
the quality of the data over time.
4Organizations should collect and curate high-quality data from r5eliable sources to ensure data
quality for their AI initiatives. High-quality data means that the data is accurate, complete,
consistent, relevant, and timely for the AI task. Reliable sources mean that the data is
trustworthy, credible, and authoritative. Collecting and curating high-quality data from reliable
sources
can improve the performance and reliability of AI systems.

The correct answer is C. Age. Age is a data quality dimension that is essential for the custom
service analytics application to analyze cases in Salesforce. Age refers to how recent the data is
and how well it reflects the current situation. For the custom service analytics application, the
data should be as fresh and up-to-date as possible, as older data may not capture the latest
status, progress, or resolution of the cases. Using outdated data can lead to inaccurate or
misleading analysis and insights, and affect the efficiency and effectiveness of the case
resolution.

The correct answer is A. Testing models with diverse datasets. Inclusivity is one of the
Salesforce’s Trusted AI Principles that aims to improve the human condition and represent the
values of all those impacted, not just the creators. Testing models with diverse datasets is an
example of this principle in practice, as it helps to ensure that the models are fair, accurate, and
representative of the diverse customer base. Testing models with diverse datasets can help to
prevent and mitigate bias, avoid harmful or discriminatory outcomes, and enhance customer
satisfaction and trust.
An example of Salesforce’s Trusted AI Principle of Inclusivity in practice is testing models with
diverse datasets. Inclusivity means that AI systems should be designed and developed with
respect for diversity and inclusion of different perspectives, backgrounds, and experiences.
Testing models with diverse datasets can help ensure that the models are fair, unbiased, and
representative of the target population or domain.
The correct answer is C. To understand and generate human language. Natural language
processing (NLP) is a branch of AI that deals with the interaction between computers and
human languages. NLP is used to understand and generate human language, such as speech,
text, or sign language. NLP can enable various AI capabilities, such as:
 Speech recognition: The ability to convert spoken words into text or commands.
 Natural language understanding: The ability to extract meaning and intent from natural
language input.
 Natural language generation: The ability to produce natural language output from data or
logic.
 Machine translation: The ability to translate natural language from one language to
another.
 Sentiment analysis: The ability to identify and extract emotions, opinions, or attitudes from
natural language.

 Text summarization: The ability to create concise and informative summaries from longer
texts.
 Chatbots: The ability to converse with humans using natural language.
Natural language processing (NLP) is used in the context of AI capabilities to understand and
generate human language. NLP can enable AI systems to interact with humans using natural
language, such as speech or text. NLP can also enable AI systems to analyze and extract
information from natural
language data, such as documents, emails, or social media posts.

The correct answer is C. Consistency. Consistency is a data quality dimension that measures
how well the data values conform to a standard format, rule, or definition. Consistency is
affected by multiple variations of state and country values in contact records, as they can cause
confusion, ambiguity, or errors in data analysis and processing. For example, if some records
use the full name of a state, while others use the abbreviation, or if some records use the
country code, while others use the country name, the data is not consistent and may lead to
inaccurate or incomplete results. To improve data consistency, Cloud Kicks should use a
common format, rule, or definition for state and country values, and apply data cleansing and
validation techniques to ensure that the data values are standardized and harmonized.
Consistency is the data quality dimension that is affected by multiple variations of state and
country values in contact records. Consistency means that the data values are uniform and
follow a common standard or format across different records, fields, or sources. Inconsistent
data can cause confusion, errors, or
duplication in data analysis and processing

The correct answer is A. Potential for human bias in machine learning algorithms and the lack of
transparency in AI decision-making processes. AI development involves various ethical
challenges, such as:
 Human bias: Human bias can affect the AI development process, such as data collection,
data labeling, model selection, model training, and model evaluation. Human bias can introduce
unfairness, discrimination, or harm to the AI outcomes, and affect the quality and accuracy of
the AI solutions. For example, if the data used to train an AI model is not representative of the
diverse customer base, the model may produce biased predictions or recommendations that
favor or disadvantage certain groups of customers.
 Transparency: Transparency refers to the ability to understand how the AI works, what
data it uses, and what outcomes it produces. Transparency is important for building
trust and accountability in AI, and for ensuring that the AI is
aligned with ethical principles and values. However, transparency can be challenging to achieve,
especially for complex or black-box AI systems, such as deep neural networks, that are difficult
to explain or interpret. For example, if an AI system makes a decision that affects a customer’s
life, such as approving or rejecting a loan application, the customer may want to know how and
why the decision was made, and what factors influenced it.
Some of the ethical challenges associated with AI development are the potential for human bias
in machine learning algorithms and the lack of transparency in AI decision-making processes.
Human bias can arise from the data used to train the models, the design choices made by the
developers, or the interpretation of the results by the users. Lack of transparency can make it
difficult to understand how and why AI systems make certain decisions, which can affect trust,
accountability, and fairness.

The correct answer is A. Enhanced accuracy and reliability of medical predictions and diagnoses.
Data quality plays a primary role in this AI application, as it affects the performance and
accuracy of the algorithm that analyzes patient data and assists in medical diagnosis. Data
quality refers to the characteristics of the data, such as completeness, correctness, consistency,
timeliness, and relevance. High-quality data can help the algorithm to learn from reliable and
relevant information, and produce accurate and reliable medical predictions and diagnoses.
Low-quality data can impair the algorithm’s ability to learn effectively, and lead to inaccurate or
unreliable medical predictions and diagnoses, which can have serious consequences for the
patients’ health and well-being.
Data quality plays a crucial role in enhancing the accuracy and reliability of medical predictions
and diagnoses. Poor data quality can lead to inaccurate or misleading results, which can have
serious consequences for patients’ health and well-being. Therefore, it is important to ensure
that the data used for AI applications in healthcare is accurate, complete, consistent, and
relevant.

The correct answer is B. Determine data availability. Data availability is the first step that Cloud
Kicks should take to prepare its data for use with AI. Data availability refers to the quantity and
quality of the data that the company has access to and can use for its AI purposes. Data
availability can affect the feasibility and effectiveness of the AI solutions, as well as the time and
cost required to implement them. To determine data availability, Cloud Kicks
should:

 Identify the data sources and types that are relevant and useful for its AI goals, such as
customer data, product data, sales data, etc.
 Assess the data quality and quantity, such as completeness, correctness, consistency,
timeliness, and relevance, and identify any data gaps or issues that need to be addressed.
 Evaluate the data accessibility and usability, such as data format, data integration, data
security, data privacy, and data governance, and ensure that the data can be easily and safely
accessed and used by the AI systems.
Before using AI to optimize business operations, the company should first assess the availability
and quality of its data. Data is the fuel for AI, and without sufficient and relevant data, AI cannot
produce accurate and reliable results. Therefore, the company should identify what data it has,
where it is stored, how it is accessed, and how it is maintained. This will help the company
understand the feasibility and scope of its AI projects.
The correct answer is B. Predictive scoring, forecasting, recommendations. These are three
commonly used examples of AI in CRM, as they can help to improve sales performance and
customer satisfaction. Here is a brief explanation of each example:
 Predictive scoring: Predictive scoring is the use of AI to assign a numerical score to each
lead or opportunity based on their likelihood of conversion or closing. Predictive scoring can
help to prioritize the most promising leads or opportunities, optimize the sales funnel, and
increase the sales efficiency and effectiveness.
 Forecasting: Forecasting is the use of AI to project the future sales revenue or demand
based on historical data and current trends. Forecasting can help to set realistic and attainable
sales goals, plan and allocate resources, and identify and mitigate risks.
 Recommendations: Recommendations are the use of AI to suggest the best actions,
products, or services for each customer or prospect based on their preferences, behavior,
and needs. Recommendations can help to personalize the

customer experience, increase customer engagement and loyalty, and cross-sell or upsell
products or services
Predictive scoring, forecasting, and recommendations are three commonly used examples of AI
in CRM. Predictive scoring can help prioritize leads, opportunities, and customers based on their
likelihood to convert, churn, or buy. Forecasting can help predict future sales, revenue, or
demand based on historical data and trends. Recommendations can help suggest the best
products, services, or actions for each customer based on their preferences,
behavior, and needs.

The correct answer is C. Model accuracy. A diverse, balanced, and large dataset is a benefit for
model accuracy, as it can help to train an AI model that can generalize well to new and unseen
data. A diverse dataset contains data from various sources, domains, and perspectives, which
can help to capture the complexity and variability of the real-world data. A balanced dataset
contains data that is evenly distributed across different categories, classes, or groups, which can
help to avoid overfitting or underfitting the data. A large dataset contains data that is sufficient
and representative of the data population, which can help to reduce the noise and uncertainty
in the data. By using a diverse, balanced, and large dataset, an AI model can learn from high-
quality data and produce accurate and reliable predictions or outcomes.
Model accuracy is a benefit of a diverse, balanced, and large dataset. A diverse dataset can
capture a variety of features and patterns that are relevant for the AI task. A balanced dataset
can avoid overfitting or underfitting the model to a specific subset of data. A large dataset can
provide enough information for the
model to learn from and generalize well to new data.

"A key benefit of effective interaction between humans and AI systems is that it leads to more
informed and balanced decision making. Effective interaction means that humans and AI
systems can communicate and collaborate with each other in a clear, natural, and respectful
way. Effective interaction can help leverage the strengths and complement the weaknesses of
both humans and AI systems. Effective interaction can also help increase trust, confidence, and
satisfaction in using AI systems."

Explanation:
"Machine learning is a key characteristic of AI capabilities that uses algorithms to learn from
data and make decisions. Machine learning is a branch of AI that enables computers to learn
from data without being explicitly programmed. Machine learning algorithms can analyze data,
identify patterns, and make predictions or recommendations based on the data."

"Duplicate management should be used to remove multiple records for the same customer in
Salesforce. Duplicate management is a feature that helps prevent and manage duplicate
records in Salesforce. Duplicate management
can help define matching rules, duplicate rules, and alert messages to detect
and merge duplicate records."
"A text field type should be used to capture the customer's preferred name. A text field type
allows the user to enter any combination of letters, numbers, or symbols. A text field type can
be used to store names, addresses, phone numbers, or other personal information."

"A possible outcome of poor data quality is that biases in data can be inadvertently learned and
amplified by AI systems. Poor data quality means that the data is inaccurate, incomplete,
inconsistent, irrelevant, or outdated for the AI task. Poor data quality can affect the
performance and reliability of AI systems, as they may not have enough or correct information
to learn from or make accurate predictions. Poor data quality can also introduce or exacerbate
biases in data, such as human bias, societal bias, or confirmation bias, which can affect the
fairness and ethics of AI systems."

"Demographic data should be omitted to avoid introducing unintended bias to an AI model.

Demographic data is data that describes the characteristics of a population or a group of
people, such as age, gender, race, ethnicity, income, education, or occupation. Demographic
data can lead to bias if it is used to discriminate or treat people differently based on their
identity or attributes. Demographic data can also reflect existing biases or stereotypes in society
or
culture, which can affect the fairness and ethics of AI systems."

"AI infringes on privacy when user consent is not obtained. User consent is the permission or
agreement given by a user to allow their personal data to be collected, used, shared, or stored
by others. User consent is an important aspect of data privacy, which is the right of individuals
to control how their personal data is handled by others. AI infringes on privacy when user
consent is not obtained because it violates the user's rights and preferences regarding their
personal data."

"High-quality data improves the reliability and credibility of AI-driven decisions, fostering trust
among users. High-quality data means that the data is accurate, complete, consistent, relevant,
and timely for the AI task. High-quality data can improve the performance and reliability of AI
systems, as they have enough and correct information to learn from and make accurate
predictions. High- quality data can also improve the trustworthiness of AI-driven decisions, as
users can have more confidence and satisfaction in using AI systems."
 Review AI Fundamentals
1 Which AI type plays a crucial rôle in Salesforce’s
predictive text and speech recognition capabilities,
enabling the plateform to understand and respond to
user
commands accurately

2 Which feature of marketing cloud einstein uses AI to

predict consumer engagement with
email and mobilePuch MESSAGING

3 What is a unique and distinguishing feature of deep

learning in the context of AI

4 Salesforce consulant is duscussing AI capabilities with a

customer who is interested in improving their sales
processes . which type of AI would be most suitable for
enhancing
sales processes in salesforce consumer 360

Explore AI Capabilities in CRM

1 What are the three main types of AI capabilities in
salesforce

2 Which salesforce AI application is recommended to

enhance sales processes

3 What is the key benefit of implementing AI in CRM

System
A. Computer vision B
B. Natural language processing NLP
C. Predictive analytics

A. Content section C
B. Email recommendations
C. Engagement scoring

A. Deep learning uses neural networks with multiple layers A

to learn from a large amount of data
B. Deep learning uses historical data to predict future
outcomes
C. Deep learning uses algorithme to cleanse and
prepare data for AI implementations

A. Predictive analytics A
B. Computer vision
C. Natural Language Processing NLP

a. Predictive generative analytic a

b. Predictive reactive analytic
c. Generative descriptive analytic d.

A. Einstein predictive builder C

B. Einstein voice
C. Einstein lead scoring D.

A. enhanced customer support A

B. improved platform speed
C. reduced data governance
NLP enables computers to understand interpret and generate human language in meaningful
way

Consumer data and machine learning are used to assign scores for every contact’s likelihood to
engage with emails and interact with push notifications

Deep learning apart from other types of AI that may not use neural networks or may use them
in a diffrent way

THIS TYPE OF AI can enhance sales processes by predicting future outcomes based on historical
data

Einstein lead scoring is specifically designed to enhance sales processes by scoring leads based
on their likelihood to convert allowing sales teams to prioritize their efforts effectively
4 Cloud kicks is implementing ai in its CRM system and is
focusing on data management
what is the benefit of using a data management
approach in AI implementation ?

Examine the Ethical Considerations of AI

1 A consultant discusses the role of humans in AI-driven
CRM processes with a customer. What is one challenge
the consultant should
mention about human-AI collaboration in decision-
making?

2 Cloud Kicks wants to implement Salesforce's AI features.

They are concerned about potential ethical and privacy
challenges. What should be recommended to minimize
potential AI bias?

3 A consultant designs a new AI model for a financial

services company that offers personal loans. Which
variable within their proposed model might introduce
unintended bias?

4 Cloud Kicks is planning to automate its customer service

chat using natural language processing. According to
Salesforce's Trusted AI principles, how should this be
disclosed to the customer?

Dig Into Data for AI

1 Cloud Kicks wants to implement AI features within its
CRM system. They have expressed concerns about the
quality of their existing
data. What advice should be given to them
 A. eliminates the need for data governance c
B. reduces the amount of data in the CRM system
C. emphasizes the importance of data quality

A Difficulty in interpreting AI decisions A

B High cost of AI implementation C
Lack of technical skills in the team

A Salesforce's Trusted AI Principles A

B Demographic data to identify minority groups
C AI models that auto-correct biased data

A Loan Date B
B Postal Code
C Payment Due Date

A They do not need to be informed they are chatting with AI. C

B Inform the customer that they are chatting with AI when they
request a live agent.
C Inform them at the beginning of the interaction that they are
chatting with AI.

A Assessing data quality is only necessary for large datasets. C

B AI systems can handle any data inaccuracies.
Data quality preparation and cleansing and data governance are all essential when
implementing AI

AI decisions are often based on complex algo and large datasets making them difficult for
humans to interpret without sufficient expertise and anderstanding of AI principles

Correct. These principles guide the development and use of AI within Salesforce, ensuring that it
is used ethically and responsibly, which includes minimizing potential AI bias.

Correct. Postal codes can Introduce blas as they are often correlated with socioeconomic status
and race. This is due to historical practices such as redlining, where certain neighborhoods were
marked as hazardous, often denying access to low-cost home lending to minority groups
residing in these areas.

Correct. This allows customers to understand the context of their interaction and sets
appropriate expectations.

Correct. High-quality data is essential for AI systems as it directly

impacts the accuracy and reliability of AI-driven predictions and
insights.
 regarding the importance of data quality for AI
implementations?

2 What role does data play in AI models?

3 Which data quality dimension refers to the frequency

and timeliness of data updates?

4 A Salesforce consultant is considering the data sets to

use for training AI models for a project on the Customer
360 platform. What should be considered when
selecting the data sets for the AI models?
C Assessing and improving data quality is crucial for accurate AI
predictions and insights.

A Data is used for training and testing AI models. A

B Data is only used for validating AI models.
C Data is only used for testing AI models.

A Data Source B
B Data Freshness
C Data Leakage

A Duplication, accuracy, consistency, storage location and usage of C

the data sets
B Age, completeness, consistency, theme, duplication, and usage of
the data sets
C Age, completeness, accuracy, consistency,
duplication, and usage of the data sets
Correct. Training data is used to teach the AI model how to make
predictions or decisions while testing data is used to evaluate the
model's performance and accuracy.

Correct. Data freshness refers to how up to date or current the data

is, which includes the frequency and timeliness of data updates.

Correct. These are the key elements/components of data quality that

are crucial when selecting data sets for AI models.