View Full Microsoft GH-900 Exam Dumps and Practice Test Dumps

 

Question 361

What is a common purpose of AI-powered chatbots?

1. Replacing every employee
2. Removing customer communication
3. Answering routine user questions
4. Eliminating support channels

Correct Answer: 3

Explanation:

AI-powered chatbots can assist users by responding to routine questions through conversational interfaces. Organizations may use them for customer support, employee assistance, frequently asked questions, and other repetitive interactions. A chatbot can provide information quickly and may operate continuously, helping reduce the workload associated with common requests. However, chatbots may not be suitable for complex or sensitive situations. Organizations should provide escalation options when users need human assistance. The chatbot should also be evaluated regularly to ensure that its responses remain useful, accurate, and appropriate for the intended audience and business process.

Question 362

What does AI-based anomaly detection identify?

1. Unusual patterns or events
2. Standard document formatting
3. Employee job titles
4. Software licensing terms

Correct Answer: 1

Explanation:

Anomaly detection identifies patterns or events that differ from expected behavior. Organizations can use this capability to identify unusual transactions, unexpected system activity, equipment problems, or other events that may require investigation. An anomaly does not automatically indicate fraud, an attack, or an error. Instead, it signals that something differs from an established pattern. Organizations can then investigate the event and determine whether action is necessary. Effective anomaly detection depends on suitable data, appropriate thresholds, and an understanding of what normal behavior looks like within the specific business environment.

Question 363

Which AI capability can help translate spoken language?

1. File compression
2. Image segmentation
3. Database indexing
4. Speech translation

Correct Answer: 4

Explanation:

Speech translation uses AI capabilities to process spoken language and produce translated content in another language. This technology can support multilingual meetings, customer interactions, accessibility scenarios, and international communication. Performance can vary based on language support, background noise, pronunciation, speaker characteristics, and the complexity of the conversation. Important communications may require human verification because automated translation can sometimes misunderstand context or terminology. Organizations should evaluate speech translation using realistic examples from their intended environment before depending on it for sensitive, highly technical, or business-critical communications.

Question 364

Why can AI improve employee productivity?

1. By removing all employee responsibilities
2. By assisting with repetitive tasks
3. By eliminating business objectives
4. By preventing collaboration

Correct Answer: 2

Explanation:

AI can improve productivity by assisting employees with repetitive or information-intensive activities. Examples include drafting content, summarizing information, extracting details from documents, answering routine questions, and organizing information. By reducing the time spent on certain tasks, AI can allow employees to focus on activities that require judgment, creativity, and human interaction. The actual productivity benefit depends on the suitability of the AI solution and the quality of its implementation. Organizations should measure outcomes rather than assuming that simply introducing an AI tool will automatically improve employee performance.

Question 365

What is a characteristic of a well-defined AI use case?

1. It has a clear business objective
2. It avoids measurable outcomes
3. It requires no relevant data
4. It ignores potential risks

Correct Answer: 1

Explanation:

A well-defined AI use case identifies a specific business problem and establishes what the organization wants to achieve. A clear objective makes it easier to determine the required data, technology, users, controls, and success measures. For example, an organization might use AI to reduce the time required to categorize customer requests. The objective can then be measured through processing time, accuracy, or another relevant metric. Defining the use case also helps prevent organizations from adopting AI simply because the technology is available without identifying a meaningful business need.

Question 366

What can Azure AI services provide to developers?

1. Physical office equipment
2. Managed AI capabilities
3. Employee payroll systems
4. Manual accounting services

Correct Answer: 2

Explanation:

Azure AI services provide managed AI capabilities that developers can integrate into applications. Depending on the service, these capabilities can support language processing, speech, computer vision, document analysis, search, and other scenarios. Using managed services can reduce the need to build and maintain every AI capability independently. Developers can instead focus on integrating appropriate functionality into their applications. Organizations should still evaluate factors such as security, privacy, cost, performance, compliance, and data requirements before selecting a particular service for a business workload.

Question 367

What does responsible AI transparency help communicate?

1. Employee private passwords
2. Confidential customer records
3. System limitations and intended use
4. Internal security credentials

Correct Answer: 3

Explanation:

Transparency can help users understand an AI system’s intended purpose, capabilities, limitations, and appropriate usage. Providing this information helps users develop realistic expectations and recognize situations where additional review may be necessary. Transparency does not require organizations to expose confidential information or sensitive security details. Instead, organizations should provide relevant information that helps stakeholders understand how the system should be used. Clear documentation and communication can support responsible AI adoption by explaining what an AI solution can do, where it may be unreliable, and what responsibilities remain with human users.

Question 368

What can AI help identify in customer-service conversations?

1. Relevant topics and intents
2. Employee salaries
3. Hardware specifications
4. Building maintenance schedules

Correct Answer: 1

Explanation:

AI can analyze customer-service conversations to identify topics, intents, sentiment, and recurring issues. For example, an AI system may categorize a conversation as a billing question, technical support request, account issue, or another predefined category. This information can help organizations route cases and identify common customer concerns. Automated language analysis can sometimes misunderstand ambiguous expressions or context. Therefore, organizations should evaluate accuracy and establish suitable review or escalation processes. AI analysis can support customer-service employees without necessarily replacing their judgment or responsibility for important customer interactions.

Question 369

Why is representative training data important?

1. It guarantees perfect predictions
2. It can support reliable model behavior
3. It removes the need for testing
4. It prevents every possible bias

Correct Answer: 2

Explanation:

Representative training data helps an AI model learn from examples that reflect the situations it is expected to encounter. If important groups, conditions, or scenarios are missing from the data, model performance may vary across different real-world situations. Representative data can therefore improve the usefulness and reliability of an AI solution. It does not guarantee perfect predictions or eliminate every form of bias. Organizations should combine suitable data practices with testing, evaluation, monitoring, and responsible AI processes to understand how the system behaves after deployment.

Question 370

What is a potential use of AI in document management?

1. Replacing office furniture
2. Removing document ownership
3. Physically filing every paper
4. Automatically classifying documents

Correct Answer: 4

Explanation:

AI can assist with document management by automatically classifying documents according to their content or business purpose. For example, invoices, contracts, applications, and reports can be categorized based on information contained within them. This can reduce manual sorting and improve processing efficiency. Organizations should evaluate classification accuracy using representative documents because document structures and terminology can vary. Security and privacy should also be considered when AI processes business records. Appropriate access permissions, retention policies, and validation processes should remain in place even when classification is automated.

Question 371

Which approach can help reduce errors in AI-supported workflows?

1. Combining AI with human review
2. Removing all validation steps
3. Accepting every output automatically
4. Disabling monitoring

Correct Answer: 1

Explanation:

Human review can help identify inaccurate or inappropriate AI outputs before they affect a business process. This is particularly useful when AI handles ambiguous information or when incorrect results could have significant consequences. Human reviewers should have sufficient context, training, and authority to evaluate outputs effectively. Organizations can also combine human review with automated validation, testing, monitoring, and escalation procedures. The required level of oversight depends on the use case and risk level. AI can provide valuable assistance while humans retain responsibility for appropriate decisions and actions.

Question 372

What can generative AI create from suitable instructions?

1. Physical network cables
2. New text or other content
3. Computer processors
4. Office access cards

Correct Answer: 2

Explanation:

Generative AI can create new content based on prompts, context, and model capabilities. Depending on the system, generated content can include text, images, audio, code, or other formats. Organizations can use these capabilities for drafting communications, summarizing information, generating ideas, creating content, or assisting with software development. Generated content should be reviewed because AI systems can produce inaccurate or unsuitable information. Clear instructions and relevant context can improve the usefulness of results, but users should still apply appropriate judgment and follow organizational policies when using generated content.

Question 373

What is an important consideration when deploying an AI application?

1. Office wall color
2. Keyboard manufacturer
3. Security and privacy requirements
4. Employee desk position

Correct Answer: 3

Explanation:

Security and privacy should be considered when deploying AI applications because these systems may process sensitive business or personal information. Organizations may need authentication, authorization, encryption, monitoring, and appropriate data-handling controls. Privacy considerations can include what information is collected, why it is processed, who can access it, and how long it is retained. Requirements vary according to the application and applicable obligations. Deployment planning should therefore consider both technical and organizational safeguards instead of focusing exclusively on model functionality or performance.

Question 374

Which capability helps AI systems understand images?

1. Computer vision
2. Speech recognition
3. Text summarization
4. Language translation

Correct Answer: 1

Explanation:

Computer vision enables AI systems to analyze visual information such as images and video. Depending on the solution, computer vision can support object detection, image classification, optical character recognition, and other visual-analysis tasks. Organizations can use these capabilities for manufacturing inspection, retail applications, document processing, security, and quality control. Performance may vary depending on image quality, lighting, viewpoint, and other environmental conditions. Testing should therefore use examples that represent the actual operating environment. Human review may also be appropriate when incorrect visual interpretation could create significant consequences.

Question 375

What can AI-powered search help employees find?

1. Physical office equipment
2. Relevant information within large collections
3. Employee attendance locations
4. Network cable lengths

Correct Answer: 2

Explanation:

AI-powered search can help users locate relevant information within large collections of documents, messages, and other content. AI techniques may interpret concepts, language, context, and user intent to provide more relevant results than simple keyword matching. This can help employees locate useful information more efficiently. However, search results may contain outdated, incomplete, or incorrect material. Users should verify important information before relying on it. Appropriate permissions must also remain in place so that search capabilities do not expose content to people who are not authorized to access it.

Question 376

What does AI scalability generally refer to?

1. Reducing all system capabilities
2. Removing application users
3. Handling increased workloads or usage
4. Limiting every process to one task

Correct Answer: 3

Explanation:

AI scalability refers to the ability of an AI system or service to handle changing workloads, users, or processing requirements. Scalability becomes important when an organization expands an AI capability to additional departments or experiences significant increases in usage. Cloud services can provide resources designed to accommodate changing workloads, although capacity limits and costs still need to be managed. Organizations should evaluate expected demand, performance requirements, service limitations, and pricing. Scalability should also be considered alongside reliability, security, and operational requirements when planning an AI deployment.

Question 377

What should users do when an AI output appears unreliable?

1. Verify the information before using it
2. Publish it immediately
3. Remove all source data
4. Disable security controls

Correct Answer: 1

Explanation:

Users should verify AI-generated information when an output appears unreliable or when the information is important. Verification can involve checking trusted sources, reviewing original documents, consulting subject-matter experts, or comparing the result against established business information. The appropriate verification method depends on the task and its potential consequences. Users should not assume that confident or fluent AI-generated content is automatically accurate. Organizations can also establish procedures for reporting problematic outputs and improving AI workflows when recurring issues are identified.

Question 378

Which factor can affect the performance of a machine learning model?

1. Desk location
2. Monitor size
3. Data quality
4. Keyboard shape

Correct Answer: 3

Explanation:

Data quality can significantly affect machine learning performance. Problems such as missing values, inaccurate information, duplicates, inconsistent formats, or poor representation can influence what a model learns and how well it performs. Organizations should therefore evaluate and prepare data carefully before using it for model development. Data quality is only one factor involved in model performance. Model design, training methods, evaluation procedures, and operating conditions can also have significant effects. A strong AI development process considers these factors together rather than assuming that model performance depends on a single element.

Question 379

Why can human judgment remain important when using AI?

1. AI always knows business priorities
2. AI may lack complete context
3. AI guarantees correct decisions
4. AI eliminates uncertainty

Correct Answer: 2

Explanation:

Human judgment can remain important because an AI system may not have access to every relevant detail surrounding a situation. A model may identify patterns or provide a recommendation while missing business priorities, contextual information, or unusual circumstances. Humans can evaluate additional information and determine whether an AI-generated recommendation is appropriate. The required level of human involvement depends on the use case and potential consequences. Human oversight is especially relevant when AI outputs influence complex, sensitive, or high-impact activities where an incorrect result could create meaningful business or customer consequences.

Question 380

Why should an AI solution’s intended purpose be documented?

1. To prevent all future updates
2. To remove user responsibility
3. To guarantee perfect outputs
4. To clarify appropriate usage

Correct Answer: 4

Explanation:

Documenting an AI solution’s intended purpose helps users and administrators understand what the system is designed to accomplish and where it should be used. Documentation can describe supported scenarios, limitations, expected inputs, known risks, and user responsibilities. Clear documentation can reduce inappropriate use and help organizations determine whether the solution remains suitable as business requirements change. It can also support training, governance, troubleshooting, and auditing. Because AI systems and business environments evolve, documentation should be reviewed and updated when capabilities, risks, policies, or intended uses change.