AI Associate Ethical Considerations of AI Practice Test 8 — 15 Questions

Question 1

A company uses Einstein's predictive lead scoring. The model inadvertently overweights leads from certain geographic regions. Which action aligns with Salesforce's Responsible AI principles?

Accepted Answer

Implement a feedback loop to continuously monitor and adjust.. Option A is correct because Salesforce's Responsible AI principles emphasize ongoing human oversight, transparency, and accountability, so implementing a feedback loop to continuously monitor and adjust the model directly addresses detected bias while keeping humans in control. This approach aligns with the Einstein Trust Layer's guidance to regularly review model outputs, measure fairness, and remediate drift or skewed weighting rather than ignoring it. Option B does not fit because simply removing the geographic feature can degrade model performance and does not by itself establish the monitoring and governance Responsible AI requires. Option C is wrong because accepting a biased model just for overall accuracy violates fairness and accountability principles, and Option D is wrong because restricting use to favorable regions hides rather than mitigates the bias.

Answer

Remove the geographic feature from the model.

Answer

Use the model as-is because it improves overall accuracy.

Answer

Only use the model for regions where it performs well.

Question 2

Which TWO practices contribute to ethical AI transparency?

Accepted Answer

Providing user-facing explanations.. Option B is correct because providing user-facing explanations directly supports transparency: it lets affected users understand what an AI system did and why, which is a core requirement of ethical AI disclosure (e.g., GDPR Articles 13-15 rights to meaningful information about automated decision-making). Option E is correct because documenting model decisions creates an auditable record of inputs, logic, and outcomes, enabling accountability, review, and reproducibility — a foundational transparency practice in AI governance frameworks. Option A is not inherently a transparency practice: open-source algorithms may aid scrutiny, but transparency concerns disclosure of system behavior and decisions, not merely code availability, and proprietary models can still be transparent. Option C is not a transparency practice and can increase privacy and bias risks, since collecting sensitive demographic data requires justification and safeguards rather than contributing to transparency. Option D concerns user control and consent (autonomy), not transparency; opting out does not by itself make an AI system's operations or decisions more understandable.

Answer

Using open-source algorithms.

Answer

Collecting sensitive demographic data.

Answer

Allowing users to opt out of AI processing.

Question 3

Which THREE actions align with Salesforce's responsible AI principles?

Accepted Answer

Involving diverse stakeholders in AI development.. Option A is correct because Salesforce's responsible AI principles emphasize inclusive, multi-stakeholder development so that diverse perspectives, values, and potential harms are surfaced early in the AI lifecycle. Option B is correct because human oversight for critical decisions is a core tenet of responsible AI, ensuring that consequential outcomes remain accountable to people rather than being fully delegated to automated systems. Option C is correct because continuous monitoring of model performance for bias is required to detect drift, disparate impact, and degradation after deployment, aligning with Salesforce's commitment to ongoing testing and evaluation. Option D is not correct because responsible AI does not mandate using only internal data sources; external or third-party data can be used provided it is lawfully sourced, representative, and governed appropriately. Option E is not correct because prioritizing accuracy over fairness inverts the responsible AI principle that fairness, safety, and trust must be balanced with, not subordinated to, raw performance metrics.

Answer

Using only internal data sources.

Answer

Prioritizing model accuracy over fairness.

Question 4

Which TWO are best practices for mitigating bias in AI models?

Accepted Answer

Using balanced training datasets.. Option B is correct because using balanced training datasets ensures that the data used to train the AI model proportionally represents all relevant groups, which directly reduces the risk of the model learning and amplifying skewed patterns or underrepresenting certain populations. Option E is correct because applying disparity analysis involves measuring outcomes across different demographic groups (e.g., via metrics like demographic parity, equalized odds, or disparate impact ratios) to detect and quantify bias, which is a fundamental step in identifying and mitigating unfair model behavior. Option A is not a best practice for bias mitigation because complex deep learning models can increase opacity and make bias harder to detect, and model complexity itself does not address data or algorithmic fairness. Option C is not correct because simply removing sensitive attributes does not eliminate bias, since proxy variables and correlated features can still encode protected information. Option D is not correct because training only on the most recent data can introduce temporal bias, exclude historically underrepresented groups, and does not by itself ensure fairness across populations.

Answer

Using complex deep learning models.

Answer

Removing all sensitive attributes.

Answer

Training models on the most recent data only.

Question 5

Refer to the exhibit. An administrator runs an audit on a sentiment analysis model. What is the primary ethical concern?

Accepted Answer

The model has low accuracy on negative text, indicating potential bias.. The correct answer is C: the model has low accuracy on negative text, indicating potential bias. In a sentiment analysis audit, a marked disparity in performance across classes—here, degraded accuracy on negative samples—is the primary ethical concern because it means the model systematically misclassifies or under-serves a specific group of inputs, producing unfair or harmful outcomes. Option A is wrong because a bias score below threshold does not by itself establish that no action is needed, and it ignores the observed class-specific accuracy gap. Option B is not the primary ethical concern here; imbalanced training data is a possible root cause, but the exhibit's ethical issue is the resulting biased performance. Option D is incorrect because overfitting to positive text would typically show high training accuracy and poor generalization, not specifically low accuracy on negative text as the ethical red flag.

Answer

The bias score is below threshold so no action needed.

Answer

The training data is imbalanced.

Answer

The model is overfitting to positive text.

Question 6

Refer to the exhibit. A developer receives this fairness check error. What is the most likely cause?

Accepted Answer

The model has higher false positive and false negative rates for Group B.. The fairness check error indicates a disparity in error rates between groups, and option A correctly identifies that the model has higher false positive and false negative rates for Group B, which directly violates fairness metrics such as equalized odds or equal opportunity. This scenario typically arises when a model's performance systematically differs across demographic groups, triggering fairness alerts. Option B is not the most likely cause because insufficient training data for Group A would more likely affect Group A's performance, not Group B's error rates. Option C is too generic, as overfitting alone does not specifically explain a group-based fairness violation. Option D is incorrect because the presence of a reweighting recommendation implies the model is not yet fair, and reweighting is a mitigation technique, not evidence of fairness.

Answer

The error is due to insufficient training data for Group A.

Answer

The model is overfitting.

Answer

The recommendation suggests reweighting, so the model is already fair.

Question 7

A company uses Salesforce Einstein to build an AI model that predicts customer churn. The model is trained on historical data from the past two years. During testing, the model shows significantly higher accuracy for male customers compared to female customers. What is the most ethical course of action?

Accepted Answer

Investigate the cause of the disparity, retrain the model with more representative data, and re-evaluate fairness.. The correct option is C: investigate the cause of the disparity, retrain the model with more representative data, and re-evaluate fairness. In this scenario, the model's significantly lower accuracy for female customers indicates a likely bias in the training data or feature selection, so the ethical response is to diagnose the root cause, improve data representativeness, and validate fairness metrics before deployment. Option A is insufficient because a disclaimer does not remediate discriminatory performance, and deploying a known biased model can cause harm. Option B ignores the ethical and business risk of deploying a model with disparate accuracy across protected groups. Option D is also wrong because manually forcing equal outputs is a post-hoc manipulation that can distort predictions and does not address the underlying bias.

Answer

Deploy the model but add a disclaimer that it may be less accurate for female customers.

Answer

Deploy the model as is, because it still meets the overall accuracy threshold.

Answer

Manually adjust the model's output to ensure equal churn predictions across genders.

Question 8

A retail company wants to use Einstein AI to personalize marketing offers. They plan to include customer purchase history and demographic data. What is the essential first step to ensure ethical use of customer data?

Accepted Answer

Obtain explicit consent from customers for using their data in AI personalization.. The correct answer is B: Obtain explicit consent from customers for using their data in AI personalization. Ethical AI use in marketing requires that customers knowingly agree to how their purchase history and demographic data will be used, since consent is the foundational principle for lawful and transparent data processing under privacy regulations like GDPR and CCPA. Option A is insufficient because not opting out is not the same as affirmative permission for AI-driven personalization, and opt-out frameworks still may not satisfy consent requirements. Option C is wrong because data collected for general business purposes cannot automatically be repurposed for AI personalization without a lawful basis. Option D is incorrect because anonymization, while a privacy safeguard, does not replace the need for consent and may also undermine personalization that depends on identifiable customer data.

Answer

Only use data from customers who have not opted out of marketing.

Answer

Proceed with available data since it was collected for business purposes.

Answer

Anonymize all customer data before feeding it into the model.

Question 9

A financial institution deploys an AI model to approve loan applications. The model uses features like income, credit score, and postal code. An audit reveals that the model denies loans at a higher rate for applicants in certain postal codes, which correlate with minority neighborhoods. What should the institution do to align with ethical AI principles?

Accepted Answer

Remove postal code from the model and retrain using only non-biased features.. The best answer is C, but removing postal code alone does not guarantee fairness. Postal code can act as a proxy for race or minority neighborhoods, so dropping it is a reasonable first mitigation step. However, other features such as income or credit score may still correlate with protected characteristics and continue to produce disparate impact. A proper ethical-AI response therefore requires retraining without the problematic feature AND validating the retrained model for disparate impact, using fairness metrics and ongoing monitoring. Option A is insufficient because monitoring alone does not stop ongoing discrimination, and monthly review still allows unfair denials. Option B only adds post-hoc transparency and does not fix the biased model. Option D accepts known disparate impact as a business risk, which conflicts with ethical AI obligations to prevent discrimination.

Answer

Continue using the model but monitor the denial rates monthly.

Answer

Add an explanation to applicants in affected areas about why they were denied.

Answer

Publish a report on the model's disparate impact and accept it as a business risk.

Question 10

A healthcare provider uses Einstein's Prediction Builder to predict patient readmission risk. The model outputs a risk score, but clinicians do not understand how the score is calculated. According to ethical AI principles, what should the provider implement?

Accepted Answer

Provide model interpretability by documenting and surfacing the key factors that influence each prediction, with clinician-facing explanations.. The correct answer is A: providing model interpretability by documenting and surfacing the key factors influencing predictions. Ethical AI principles emphasize transparency and interpretability, especially in healthcare, so clinicians must be able to understand which patient attributes (e.g., prior admissions, medication history) drove a given readmission risk score. Einstein Prediction Builder does not offer a native per-prediction explainability feature, so the provider should implement model interpretability through feature importance analysis and clinician-facing explanations. Option B is wrong because accuracy alone does not satisfy the ethical duty to explain clinical decision support outputs. Option C is insufficient because transparency must reach the clinicians using the score, not just the data science team. Option D is not required, since interpretability can be addressed without necessarily sacrificing accuracy.

Answer

Proceed with the model since it provides accurate predictions.

Answer

Train only the data science team on the model's inner workings.

Answer

Replace the model with a simpler, less accurate but fully transparent model.

Question 11

A company uses an AI chatbot that automatically responds to customer service inquiries. When a customer questions the bot's response, there is no mechanism for the customer to appeal or speak to a human. What ethical principle is being violated?

Accepted Answer

Accountability. Accountability, because the scenario describes a system where no mechanism exists for a customer to appeal an AI decision or escalate to a human, meaning no party is answerable for the chatbot's responses. Accountability in AI ethics requires that there be identifiable responsibility and recourse channels for automated decisions that affect users. Transparency (B) concerns disclosing how the system works and that AI is being used, which is not the issue here. Privacy (A) relates to handling of personal data, and Fairness (D) concerns bias and equitable treatment, neither of which is described in this scenario.

Answer

Privacy

Answer

Transparency

Answer

Fairness

Question 12

A company is building an AI model to score sales leads. They have a dataset with historical leads, including whether they converted. The dataset contains 90% male and 10% female leads. The model will be used to prioritize leads for sales follow-ups. What is the primary ethical concern?

Accepted Answer

The training data is imbalanced, which may cause the model to be less accurate for female leads, leading to unfair prioritization.. The correct answer is A: the training data is imbalanced, with 90% male and 10% female leads, which can cause the model to learn patterns that favor the majority class and perform less accurately for female leads, resulting in unfair prioritization in sales follow-ups. This is a classic class-imbalance and fairness concern in machine learning, where underrepresentation of a group can lead to disparate model performance and discriminatory outcomes. Option B is incorrect because overrepresentation of male leads typically biases the model in favor of, not against, the majority group. Option C is not necessarily true, as dataset size alone is not the primary ethical issue here, and no size threshold is given. Option D is too broad and not the primary concern, since using historical data is standard practice and only becomes unethical if it encodes bias or outdated discrimination.

Answer

The model will be biased against male leads because they are overrepresented.

Answer

The dataset is too small to build a reliable model.

Answer

Using historical data is unethical because it may not reflect current conditions.

Question 13

A company deploys an Einstein AI model that recommends products to customers. To ensure transparency, what should the company include in the customer-facing interface?

Accepted Answer

A brief note that recommendations are generated by AI, along with a way to get more information about why a product was suggested.. Option D is correct because transparency in AI-driven recommendations requires disclosing that the recommendations are AI-generated and providing an accessible path for customers to understand the reasoning behind a specific suggestion, such as an explanation link or 'Why am I seeing this?' feature. This aligns with responsible AI principles like explainability and user trust, ensuring customers are informed without overwhelming them with technical detail. Option A is incorrect because falsely claiming human selection misrepresents the AI system and undermines transparency. Option B is insufficient because a bare 'AI-generated' label lacks the explanatory context customers need to understand recommendations. Option C is impractical and not user-friendly, as exposing the complete feature list would overwhelm customers and may reveal proprietary or sensitive model details rather than a meaningful explanation.

Answer

A statement that each recommendation is personally selected by a human expert.

Answer

A simple label 'AI-generated' without further explanation.

Answer

A complete list of all features used by the model.

Question 14

A company wants to use Einstein OCR to extract text from uploaded documents. To protect customer privacy, what should they ensure before processing documents containing personal data?

Accepted Answer

Enable the Einstein Trust Layer to mask sensitive data and comply with data protection policies.. The correct answer is D: Enable the Einstein Trust Layer to mask sensitive data and comply with data protection policies. Einstein Trust Layer provides built-in data masking, zero-retention, and audit capabilities that prevent personal data from being exposed to or retained by the LLM during Einstein OCR processing, which directly addresses the privacy requirement. Option A is insufficient because it shifts the burden to customers and does not enforce protection at the platform level. Option B only secures data in transit via TLS/HTTPS and does nothing about masking or handling sensitive data once received. Option C is incorrect because relying on the model to ignore sensitive data is not a reliable safeguard and consent alone does not satisfy data protection obligations.

Answer

Ask customers to manually blur sensitive information before uploading.

Answer

Use HTTPS to securely upload documents.

Answer

Obtain consent from customers and rely on the model to ignore sensitive data.

Question 15

A hospital uses an AI model to predict patient deterioration. The model was trained on data from a single hospital with a predominantly white patient population. When deployed at a hospital serving a diverse population, the model underperforms for minority groups. What is the most effective way to address this ethical issue?

Accepted Answer

Retrain the model using a more diverse dataset that represents the target population.. Retrain the model using a more diverse dataset that represents the target population. The root cause of the underperformance is that the training data came from a single hospital with a predominantly white patient population, so the model learned patterns that do not generalize to minority groups; retraining on data that reflects the target population directly addresses this representativeness problem and is the most effective ethical remedy. Option A is problematic because separate models per demographic group can be discriminatory and may not be feasible for many groups, and it does not fix the underlying data gap. Option B is useful for detecting and reporting disparities but is only a monitoring/reporting measure, not a fix for the biased model. Option D merely adjusts a threshold to change sensitivity, which can mask rather than correct the underlying bias and may introduce other harms.

Answer

Create separate models for each demographic group to ensure accuracy.

Answer

Continuously monitor model performance across demographic groups and report disparities.

Answer

Adjust the decision threshold for minority groups to improve sensitivity.