ISO/IEC TR 24028:2020 - AI Trustworthiness Foundation Free Practice Test — 30 Questions

Question 1

A consortium developing an advanced AI-driven medical diagnostic tool is facing scrutiny regarding its potential for discriminatory outcomes and its susceptibility to subtle data manipulations that could lead to misdiagnoses. They are seeking to establish a robust framework for ensuring the AI\'s reliability and ethical deployment, aligning with international standards for AI trustworthiness. Which of the following AI trustworthiness attributes, as conceptualized within foundational frameworks like ISO/IEC TR 24028:2020, is most critical as a prerequisite for effectively addressing both the concerns of discriminatory outcomes and the potential for subtle data manipulations, thereby fostering overall confidence in the system\'s integrity?

Accepted Answer

Explainability

Question 2

Considering the evolving landscape of AI regulation, such as the principles embedded within the EU AI Act, which of the following best describes the relationship between regulatory compliance and the foundational pillars of AI trustworthiness as conceptualized in ISO/IEC TR 24028:2020?

Accepted Answer

Regulatory mandates serve as concrete, actionable implementations of abstract AI trustworthiness principles, requiring proactive integration into the AI lifecycle to ensure genuine confidence.

Question 3

When assessing the trustworthiness of an AI system intended for critical infrastructure management, which of the following approaches most comprehensively aligns with the principles advocated in ISO/IEC TR 24028:2020, ensuring sustained reliability and societal acceptance?

Accepted Answer

Implementing rigorous, continuous monitoring of system performance against predefined benchmarks, coupled with a transparent incident reporting mechanism and a dedicated team for ethical oversight and bias mitigation throughout the AI lifecycle.

Question 4

Consider an advanced AI system deployed for autonomous vehicle navigation in a densely populated urban environment. Recent incidents have highlighted potential edge cases where the system\'s decision-making process led to unexpected and suboptimal outcomes, raising concerns about its reliability and safety. To enhance the trustworthiness of this system in accordance with the principles of ISO/IEC TR 24028:2020, which fundamental aspect of AI trustworthiness requires the most immediate and rigorous attention to establish clear responsibility and enable effective recourse?

Accepted Answer

Establishing clear lines of human accountability and implementing robust oversight mechanisms for the AI's operational behavior.

Question 5

A multinational financial institution is developing an AI-driven credit scoring system. Given the stringent regulatory environment and the potential for discriminatory outcomes, what foundational principle from ISO/IEC TR 24028:2020 should guide the institution\'s primary focus during the system\'s validation phase to ensure both efficacy and ethical compliance?

Accepted Answer

Demonstrating that the AI system's decision-making processes are auditable and explainable to relevant stakeholders, allowing for the identification and mitigation of potential biases.

Question 6

Considering the foundational principles of AI trustworthiness as detailed in ISO/IEC TR 24028:2020, which of the following best encapsulates the overarching goal of integrating multiple assurance characteristics throughout the AI lifecycle, particularly in light of evolving regulatory landscapes such as the EU AI Act\'s risk-based approach?

Accepted Answer

Establishing a comprehensive framework for AI system assurance that systematically addresses robustness, explainability, fairness, accountability, transparency, security, and privacy across all development and operational phases.

Question 7

Considering the principles outlined in ISO/IEC TR 24028:2020 for AI trustworthiness, how would a regulatory framework that employs a risk-based classification system, imposing stringent requirements for \"high-risk\" AI applications concerning data quality, algorithmic transparency, and human oversight, most effectively contribute to the overall assurance of AI trustworthiness?

Accepted Answer

By mandating specific technical and organizational measures that directly address key trustworthiness attributes like fairness, robustness, and accountability for systems deemed high-risk.

Question 8

Consider an autonomous vehicle\'s perception system designed to identify pedestrians. If this system, despite being trained on a vast dataset of clear weather conditions, can still accurately detect pedestrians in moderate rain or fog with only a slight, predictable degradation in performance, which fundamental AI trustworthiness attribute is primarily being demonstrated?

Accepted Answer

Robustness

Question 9

Consider an advanced AI system designed for diagnostic imaging analysis in a regulated medical environment. To ensure its trustworthiness according to the principles outlined in ISO/IEC TR 24028:2020, what fundamental approach should be prioritized when conducting an initial assessment of its suitability for deployment?

Accepted Answer

A comprehensive evaluation of the system's performance across multiple, diverse datasets, coupled with an examination of its decision-making processes for explainability and its susceptibility to adversarial attacks.

Question 10

Consider an AI system deployed for autonomous vehicle navigation in a complex urban environment. During its operation, the system encounters a sudden, localized fog bank that significantly reduces sensor visibility, a condition not extensively represented in its training data. Subsequently, the AI exhibits erratic steering adjustments and fails to correctly identify lane markings, leading to a near-collision. Which fundamental AI trustworthiness characteristic, as discussed in ISO/IEC TR 24028:2020, was most critically compromised in this scenario, directly impacting the system\'s overall reliability and safety?

Accepted Answer

Robustness

Question 11

Consider a scenario where a novel AI system is developed for automated medical diagnosis in a highly regulated healthcare environment. The system has demonstrated high accuracy in laboratory testing but exhibits occasional, unpredictable deviations in its diagnostic recommendations when exposed to real-world, diverse patient data. To ensure the trustworthiness of this AI system according to the principles discussed in ISO/IEC TR 24028:2020, which of the following aspects would be the most critical to address during its deployment and operational phases?

Accepted Answer

Implementing continuous, multi-stage human validation of diagnostic outputs, with clear protocols for intervention and override based on predefined risk thresholds.

Question 12

Considering the multifaceted nature of AI trustworthiness as outlined in ISO/IEC TR 24028:2020, which fundamental characteristic most directly enables the verification of an AI system\'s adherence to ethical guidelines and regulatory compliance, thereby fostering human confidence in its operations?

Accepted Answer

The system's capacity to provide comprehensible explanations for its outputs and decisions.

Question 13

Considering the evolving global regulatory environment for artificial intelligence, such as the European Union\'s AI Act, how does the guidance provided in ISO/IEC TR 24028:2020 on AI trustworthiness interact with external legal compliance requirements for AI systems?

Accepted Answer

The TR's guidance provides a foundational framework for building AI trustworthiness that must be adapted and integrated with specific external legal mandates to ensure comprehensive compliance.

Question 14

A multinational financial institution is developing an AI-powered credit scoring system intended for use across several jurisdictions with varying data privacy regulations, such as the GDPR in Europe and CCPA in California. The system must not only provide accurate credit assessments but also ensure that its decision-making processes are auditable and do not perpetuate historical biases against protected groups. Which of the following approaches best aligns with the principles of AI trustworthiness as described in ISO/IEC TR 24028:2020, considering the regulatory landscape and the need for demonstrable compliance?

Accepted Answer

Implementing a continuous assurance framework that integrates rigorous testing for bias, explainability mechanisms, and data governance protocols aligned with relevant jurisdictional requirements throughout the AI lifecycle.

Question 15

When evaluating an AI system intended for use in a sensitive financial advisory role, which combination of trustworthiness characteristics, as discussed within the foundational principles of ISO/IEC TR 24028:2020, would be most critical for ensuring compliance with financial regulations and fostering user confidence?

Accepted Answer

Robustness against data perturbations, transparency in decision-making processes, and demonstrable fairness in client recommendations.

Question 16

A consortium developing an AI-powered medical diagnostic assistant is seeking to establish a robust framework for ensuring the trustworthiness of their system before widespread deployment. They are particularly concerned with regulatory compliance, such as the proposed EU AI Act\'s requirements for high-risk AI systems, and the need to instill confidence in both medical professionals and patients. Which of the following activities most directly addresses the foundational requirement for building this confidence by proactively identifying and mitigating potential failure modes and biases that could compromise the system\'s reliability and ethical operation?

Accepted Answer

Conducting a comprehensive risk assessment and implementing a continuous monitoring program to detect and address deviations from expected performance and identify emergent biases.

Question 17

Consider an advanced AI system deployed in a critical infrastructure monitoring role. This system has undergone extensive validation and consistently provides accurate anomaly detection alerts, with its performance metrics demonstrating a low rate of false positives and false negatives across a wide range of operational scenarios. The system\'s design prioritizes predictable and correct outputs within its defined operational domain. Which primary trustworthiness attribute, as conceptualized within the framework of AI trustworthiness, is most prominently exemplified by this system\'s performance?

Accepted Answer

Reliability

Question 18

In the context of establishing AI trustworthiness according to ISO/IEC TR 24028:2020, consider an AI system designed to assist in loan application assessments. If this system consistently denies applications from a specific demographic group, what fundamental trustworthiness attribute is most directly challenged, necessitating detailed investigation into the AI\'s decision-making process to ensure compliance with principles like fairness and non-discrimination, as often mandated by financial regulations?

Accepted Answer

Explainability

Question 19

A multinational corporation is developing an AI-powered diagnostic tool for a highly regulated medical field. To ensure public trust and compliance with emerging AI governance frameworks, what is the most comprehensive approach to demonstrating the AI system\'s trustworthiness, considering the principles outlined in ISO/IEC TR 24028:2020?

Accepted Answer

Implementing a robust system of continuous monitoring, independent third-party audits, and detailed documentation of the AI's decision-making processes and performance metrics throughout its lifecycle.

Question 20

Consider an advanced autonomous drone system designed for environmental monitoring in remote, sensitive ecological zones. The system is equipped with sophisticated AI for data collection, analysis, and navigation. To ensure trustworthiness, which of the following operational principles, as outlined in foundational AI trustworthiness frameworks, would be most critical for maintaining human control and accountability in the event of unforeseen environmental changes or system anomalies?

Accepted Answer

Ensuring the system allows for immediate human intervention to alter flight paths or halt operations based on real-time visual or sensor data interpretation by a remote operator.

Question 21

Considering the evolving landscape of AI regulation, exemplified by frameworks like the EU AI Act which categorizes AI systems by risk, how can an organization most effectively demonstrate the trustworthiness of its AI-powered diagnostic imaging system, classified as high-risk, to regulatory bodies and end-users?

Accepted Answer

Implementing rigorous data validation protocols, conducting regular adversarial testing to ensure robustness against manipulation, establishing clear audit trails for decision-making processes, and providing comprehensive documentation on bias mitigation strategies.

Question 22

Consider an advanced AI system designed to manage a nation\'s critical infrastructure, such as its energy distribution network. In the context of ISO/IEC TR 24028:2020, which fundamental characteristic of AI trustworthiness is most directly addressed by ensuring the AI can consistently and accurately perform its intended functions, even when encountering novel or unexpected operational conditions, thereby preventing catastrophic failures?

Accepted Answer

Robustness and reliability

Question 23

Consider a scenario where a financial institution deploys an AI system for loan application processing. The system, while achieving high accuracy in predicting loan defaults, exhibits opaque decision-making patterns. To enhance the trustworthiness of this system in alignment with ISO/IEC TR 24028:2020 principles, which of the following actions would most effectively address the identified trustworthiness gap?

Accepted Answer

Implementing post-hoc explanation techniques to provide insights into the model's reasoning for individual loan decisions, alongside establishing a clear governance framework for AI model validation and ongoing performance monitoring.

Question 24

Consider an advanced AI system designed for medical diagnostics that has undergone extensive training on a diverse dataset. During deployment, it encounters a novel, rare genetic mutation not present in its training corpus, leading to a misdiagnosis. Which fundamental AI trustworthiness characteristic, as elaborated in ISO/IEC TR 24028:2020, is most directly challenged by this scenario, and what is the primary implication for the system\'s reliability in unforeseen circumstances?

Accepted Answer

Robustness, indicating a potential failure to maintain performance under conditions outside its training distribution.

Question 25

A financial institution deploys an AI system for automated loan application assessment. Following its implementation, a significant number of applications from a particular demographic group are consistently rejected, suggesting potential algorithmic bias. According to the principles of AI trustworthiness as detailed in ISO/IEC TR 24028:2020, which of the following actions would be most critical for the institution to undertake to address this issue and uphold responsible AI practices?

Accepted Answer

Conduct a thorough audit of the AI system's data inputs, model architecture, and decision-making logic to identify the root cause of the biased outcomes and establish a clear chain of accountability.

Question 26

Consider an advanced AI system designed for personalized medical treatment recommendations. To ensure its trustworthiness according to the principles outlined in ISO/IEC TR 24028:2020, what is the most critical element to demonstrate during its post-deployment evaluation phase, especially when considering potential regulatory scrutiny under frameworks like the EU AI Act?

Accepted Answer

Comprehensive audit trails of all data inputs, model inferences, and human-AI interaction logs, coupled with a documented process for continuous monitoring and bias mitigation.

Question 27

Consider a scenario where an advanced AI system, deployed for critical decision-making in a regulated industry, exhibits emergent behaviors that are not fully understood by its developers. This opacity leads to concerns about potential discriminatory outcomes, which could contravene established data protection and anti-discrimination legislation. The organization is seeking to enhance the system\'s trustworthiness. Which of the following strategies would most effectively address the identified issues and align with the principles of AI trustworthiness as described in foundational standards like ISO/IEC TR 24028:2020, while also considering the need for regulatory compliance?

Accepted Answer

Implement a comprehensive framework for AI explainability and interpretability, coupled with robust governance structures for continuous monitoring, bias detection, and human oversight, ensuring that the system's decision-making processes are auditable and aligned with legal requirements.

Question 28

Consider an advanced AI system designed for medical diagnosis. While the system demonstrates exceptional accuracy in identifying rare diseases, its decision-making process is entirely opaque, and it lacks any mechanism for a human clinician to review or override its recommendations. Furthermore, preliminary testing suggests a statistically significant disparity in diagnostic accuracy between demographic groups, though the exact cause of this bias is not readily apparent from the system\'s internal workings. Which of the following best reflects the primary shortcomings of this AI system concerning the principles of AI trustworthiness as described in ISO/IEC TR 24028:2020?

Accepted Answer

Insufficient human agency and oversight, coupled with a lack of transparency and potential for unfairness due to algorithmic bias.

Question 29

A multinational fintech company is developing an AI-powered credit scoring model intended for use across several European Union member states. Given the evolving landscape of AI regulation, including directives like the proposed EU AI Act and existing data protection laws such as GDPR, the company must rigorously demonstrate the trustworthiness of its model. Which of the following practices would most effectively enable the company to provide verifiable evidence of the model\'s adherence to trustworthiness principles throughout its lifecycle to regulatory bodies?

Accepted Answer

Implementing a comprehensive, auditable log of all data inputs, model parameters, training procedures, validation metrics, and performance monitoring results, linked to specific development and deployment phases.

Question 30

Consider an advanced AI system designed for critical infrastructure management. Following a significant operational anomaly that led to a temporary service disruption, an investigation reveals that while the system\'s algorithms performed within expected parameters based on its training data, the integration of a new sensor input, which was not adequately validated for its data quality, contributed to the unexpected behavior. According to the principles of AI trustworthiness as discussed in ISO/IEC TR 24028:2020, which aspect of governance is most crucial to address to prevent recurrence and enhance accountability in such a scenario?

Accepted Answer

Establishing a clear framework for assigning responsibility for data validation and integration processes throughout the AI system lifecycle.

ISO/IEC TR 24028:2020 - AI Trustworthiness Foundation Free Practice Test — 30 Questions

A lot more is already mapped out

About the ISO/IEC TR 24028:2020 - AI Trustworthiness Foundation Certification