AI, AI-900, Artificial Intelligence (AI), Generative AI, Microsoft Certification

Identify Features of Generative AI Workloads (AI-900 Exam Prep)

Overview

Generative AI is a class of Artificial Intelligence (AI) workloads that create new content rather than only analyzing or classifying existing data. On the AI-900: Microsoft Azure AI Fundamentals exam, you are expected to understand what generative AI is, what kinds of problems it solves, and how it differs from other AI workloads—not how to train large models or write code.

This topic appears under:

  • Describe Artificial Intelligence workloads and considerations (15–20%)
    • Identify features of common AI workloads

Expect conceptual and scenario-based questions that test whether you can recognize when generative AI is the appropriate approach.

What Is a Generative AI Workload?

A generative AI workload uses models that can generate new, original content based on patterns learned from large datasets.

Generative AI systems can produce:

  • Text (responses, summaries, stories, code)
  • Images (artwork, illustrations, designs)
  • Audio (music, speech)
  • Video (short clips or animations)

Key defining feature:
Unlike traditional AI that predicts or classifies, generative AI creates.

Common Generative AI Use Cases

On the AI-900 exam, generative AI is typically presented through productivity, creativity, or assistance scenarios.

Text Generation

What it does: Generates human-like text based on a prompt.

Example scenarios:

  • Drafting emails or reports
  • Writing marketing copy
  • Generating code snippets
  • Creating conversational responses

Key idea: The model produces new text rather than selecting from predefined responses.

Summarization

What it does: Creates concise summaries of longer text.

Example scenarios:

  • Summarizing documents or meeting notes
  • Condensing long articles

Exam note: Summarization may appear in both NLP and generative AI contexts. When the output is newly generated text, it is generative AI.

Question Answering and Chat Experiences

What it does: Generates natural language answers to user questions.

Example scenarios:

  • AI chat assistants
  • Knowledge-based Q&A systems

Key idea: Responses are generated dynamically rather than retrieved verbatim.

Image Generation

What it does: Creates images from text descriptions.

Example scenarios:

  • Generating illustrations or artwork
  • Creating marketing visuals

Key idea: The system produces entirely new images rather than analyzing existing ones.

Code Generation

What it does: Generates programming code from natural language prompts.

Example scenarios:

  • Creating sample scripts
  • Explaining or completing code

Azure Services Associated with Generative AI

For AI-900, service knowledge is high-level and conceptual.

Azure OpenAI Service

Supports:

  • Text generation
  • Chat-based experiences
  • Image generation
  • Code generation

This is the primary Azure service associated with generative AI workloads on the exam.

How Generative AI Differs from Other AI Workloads

Recognizing these differences is critical for AI-900.

AI Workload TypePrimary Output
Generative AINewly created content
Natural Language ProcessingAnalysis of text
Computer VisionAnalysis of images and video
Document ProcessingStructured data extraction
Speech AITranscription or audio generation

Exam tip: If the system is creating something new (text, image, code), think generative AI.

Prompt Engineering (Conceptual Awareness)

AI-900 includes basic awareness of prompting.

Prompt engineering refers to crafting inputs that guide a generative model toward better outputs.

Examples:

  • Providing context
  • Specifying tone or format
  • Giving examples in the prompt

No technical depth is required, but you should understand that outputs depend on prompts.

Responsible AI Considerations

Generative AI introduces unique risks.

Key considerations include:

  • Hallucinations (incorrect or fabricated outputs)
  • Bias in generated content
  • Harmful or inappropriate responses
  • Transparency that content is AI-generated

AI-900 tests awareness, not mitigation techniques.

Exam Tips for Identifying Generative AI Workloads

  • Look for verbs like generate, create, draft, write, summarize
  • Focus on whether the output is new content
  • Ignore implementation details and model names
  • Choose generative AI when static rules or classification are insufficient

Summary

For the AI-900 exam, you should be able to:

  • Recognize scenarios that require generative AI
  • Identify common generative AI use cases
  • Associate generative AI with Azure OpenAI Service
  • Distinguish generative AI from analytical AI workloads
  • Understand high-level responsible AI considerations

Go to the Practice Exam Questions for this topic.

Go to the AI-900 Exam Prep Hub main page.

Leave a comment