AI Associate Practice Test 12 — 15 Questions

Question 1

A non-profit organization uses Data Cloud to manage donor data from multiple sources (email campaigns, event attendance, donations). They want to use an AI model to predict future donations. The data scientist says the model needs a unified view of each donor with consistent fields. What is the first step the data architect should take in Data Cloud to enable this?

Accepted Answer

Define a Data Model that maps each source's fields to a common donor object. Option B is the correct first step. Defining a Data Model that maps fields from each source to a common donor object ensures all data conforms to a single schema, which is foundational for unified views. Option A is necessary but can be done after the model. Option C is premature without unified data. Option D is too early; cleaning at source is good but not the first step in Data Cloud.

Answer

Create a Data Stream for each source (email, events, donations)

Answer

Set up Data Actions to clean data at each source

Answer

Immediately start training the AI model on raw data from the streams

Question 2

A company is preparing customer data to train a custom AI model for sentiment analysis. Which two data preparation best practices should they follow? (Choose two.)

Accepted Answer

Ensure data is representative of all customer demographics.. Ensuring representative data and anonymizing PII are critical for model fairness and privacy. Removing all records with missing values can discard useful information; using only recent data may introduce bias; single-annotator labeling can cause subjective bias.

Answer

Use only data from the last month.

Answer

Remove all records with missing values.

Answer

Label data manually by a single annotator.

Question 3

Refer to the exhibit. An admin sees this error in the Einstein activity log. What is the most likely cause?

Accepted Answer

Data masking is configured but not applied to the sentiment analysis model.. The error indicates that sentiment analysis is not enabled, but PII is leaking, likely because data masking is not applied to the model generating the output.

Answer

Sentiment analysis is generating PII that data masking cannot hide.

Answer

The model output should not contain any text.

Answer

The Einstein Trust Layer is completely disabled.

Question 4

Which THREE are core principles in Salesforce's AI ethics framework?

Accepted Answer

Fairness. Option A is correct because privacy is a core principle. Option C is correct because fairness is a core principle. Option D is correct because transparency is a core principle. Option B is wrong while important, accuracy is not explicitly a core ethical principle in Salesforce's framework; they focus on fairness, transparency, accountability, privacy, and safety. Option E is wrong profitability is not an ethical principle.

Answer

Accuracy

Answer

Profitability

Question 5

A user asks an AI assistant to generate content that may be offensive. What should the AI do?

Accepted Answer

Refuse and explain why. The AI should refuse the request and explain why, upholding ethical standards.

Answer

Ignore the request

Answer

Generate and report the user

Answer

Generate with a warning

Question 6

Refer to the exhibit. An AI model's accuracy is shown for four demographic groups. Which group should be investigated for potential bias?

Accepted Answer

Group Beta. Group Beta has significantly lower accuracy, indicating possible bias or underperformance.

Answer

Group Alpha

Answer

Group Gamma

Answer

Group Delta

Question 7

A financial institution uses Einstein Discovery to analyze loan applications. The model denies loans at a higher rate for a particular ethnicity. The data is unbiased, but the model learned societal biases. Which action BEST aligns with ethical AI practices?

Accepted Answer

Apply fairness constraints and re-evaluate the model. Option D is correct because using fairness metrics and adjusting thresholds can mitigate learned bias. Option A is wrong as it may still discriminate. Option B is wrong because explanation alone doesn't fix bias. Option C is wrong as standard metrics may not capture fairness.

Answer

Provide an explanation to denied applicants

Answer

Use the model as is since the data is unbiased

Answer

Rely on standard performance metrics like accuracy

Question 8

A Salesforce developer is building an AI model to predict customer churn. What is the most important ethical consideration when collecting training data?

Accepted Answer

Collect only data that is necessary for the prediction. The correct answer is B because data privacy is critical; collecting only necessary data minimizes risk. Option A is wrong because historical data can contain bias. Option C is wrong because older data may be less relevant. Option D is wrong because more features can increase privacy risks and complexity.

Answer

Include as many features as possible

Answer

Focus on data from the last month for relevance

Answer

Use as much historical data as possible

Question 9

A developer notices that an AI model performs differently for different age groups. What should be done?

Accepted Answer

Investigate the cause and evaluate fairness metrics.. Option B is correct because investigating the cause and evaluating fairness metrics is the proper first step. Option A is wrong ignoring can lead to unethical outcomes. Option C is wrong retraining with more data may help but without investigation may not address root cause. Option D is wrong removing age as a feature may not eliminate bias and could reduce model accuracy.

Answer

Retrain with more data from all ages.

Answer

Remove age as a feature.

Answer

Ignore it if overall accuracy is high.

Question 10

Refer to the exhibit. Which ethical principle is most directly violated?

Accepted Answer

Privacy. Option B is correct because PII leakage violates privacy. Option A is wrong fairness is about bias. Option C is wrong transparency is about openness. Option D is wrong accountability is about responsibility, but the immediate violation is privacy.

Answer

Transparency

Answer

Accountability

Answer

Fairness

Question 11

A company is developing an AI system to screen job applicants. Which TWO practices are essential for ethical AI in hiring?

Accepted Answer

Auditing the model for bias against protected groups. Options B and D are essential: auditing for bias (B) and ensuring transparency (D). Option A is wrong because speed is not ethical. Option C is wrong as using all data may include biased data. Option E is wrong because sole reliance on AI is unethical without human oversight.

Answer

Using all available data including demographic details

Answer

Relying solely on AI for final decisions

Answer

Maximizing processing speed

Question 12

A company deploys an AI system to screen job applications. The system is found to consistently reject candidates from a particular university, even though those candidates are qualified. What is the most ethical first step?

Accepted Answer

Investigate the training data and model for bias. The correct answer is A because investigating the data and model for bias is the appropriate ethical action. Option B is wrong because ignoring the bias violates fairness. Option C is wrong because immediately increasing rejections is unethical. Option D is wrong because changing the screening criteria without analysis may not address the root cause.

Answer

Increase rejections from other universities to balance

Answer

Ignore the finding as correlation, not causation

Answer

Change the screening criteria to include more universities

Question 13

A company uses an AI model to screen job applicants. They discover the model is rejecting candidates from a certain demographic at a higher rate. Which ethical principle is most clearly violated?

Accepted Answer

Fairness. Fairness requires that AI systems do not discriminate against groups. The model's disparate impact violates fairness.

Answer

Transparency

Answer

Privacy

Answer

Accountability

Question 14

A Salesforce customer uses Einstein Sentiment Analysis to analyze customer feedback. They find the model is less accurate for non-English languages. What ethical concern does this raise?

Accepted Answer

Bias. The varying accuracy across languages indicates bias in the model, which is a fairness concern.

Answer

Accountability

Answer

Privacy

Answer

Security

Question 15

An organization uses Einstein Discovery to analyze survey data. The model reveals a correlation between age and satisfaction. What is the responsible use of this insight?

Accepted Answer

Investigate causality before action. Correlation does not imply causation; responsible use involves investigating the underlying cause before acting.

Answer

Act on the correlation immediately

Answer

Discard the result

Answer

Publish the result as-is