Before using this tool ensure you have familiarised yourself with the practical and ethical considerations in the ‘Getting starting with Blackboard AI Design Assistant‘ guide
- What is prompting?
- Overview of prompting in Blackboard AI Design Assistant
- Tips for prompt writing
- How to use it
- Example prompt: Test Question Generation
- Example prompt: Image Generation
- Example prompt: AI Conversation
- Example prompt: Rubric Generation
What is prompting for Blackboard AI Design Assistant?
In the context of Blackboard’s AI Design Assistant, ‘prompting’ refers to the process of writing a short piece of text (‘Description’) that tells the AI what kind of content you want it to generate and adjusting the general settings to focus the output. The quality of the prompt significantly affects the usefulness of the AI’s output. A clear, well-structured prompt helps the tool generate relevant, accurate, and pedagogically appropriate suggestions aligned with your teaching goals. You can input up to 2000 characters in the prompt ‘Description’.
Unlike conversational AI tools such as Microsoft’s CoPilot or OpenAI’s ChatGPT, which allow for back-and-forth refinement through ongoing interaction, Blackboard’s tools generate content in a single step based on your initial input and chosen settings. There is no opportunity to adjust the AI’s response once generated, so a clear and well-structured prompt is essential.
As with any AI-generated content, you are responsible for reviewing and editing the material before it is used with students. Always check that the output suggestions are accurate, appropriate, and aligned with your learning outcomes.
Tips for prompt writing
Tip: If you are finding it difficult to write a clear and effective prompt, try using a conversational generative AI tool to help. It may sound a little oxymoronic, but you can ask ChatGPT or Microsoft CoPilot to generate a sample prompt for a specific topic or tool in Blackboard. This can be a surprisingly effective way to clarify your thinking or get ideas for structuring your input.
- Know your objective and be specific. Start by identifying what you want the AI to generate and for what purpose. Formulate your prompt with clear, concise language and include key details or constraints. The more precise you are about the topic, format, length, or complexity of the output, the more relevant and useful the AI’s suggestion will be. For example, “Write a short assignment that introduces students to the ethical implications of AI in healthcare, suitable for a second-year undergraduate module” is more effective than “Write an assignment about AI and ethics.”
- Provide context or content examples. If possible, ground your prompt in your course context. Mention specific chapters, materials, or learning outcomes you want the AI to consider. You can also specify the tone or format by giving an example. Including relevant context terms in the prompt nudges the AI to produce more course-aligned results. For instance, reference a course handout (“Based on the Week 3 handout on ethical frameworks, …”) or desired style (“in a formal academic tone”) if it is important. One-shot prompting (providing one example) or few-shot prompting (multiple examples) can demonstrate the pattern you want the AI to follow.
- Use the available tool settings to shape the output. The AI Design Assistant interface provides controls to help guide the output. Use the complexity level slider to match the academic level and depth you want for your students. Higher levels produce more advanced vocabulary or depth, for example Level 7 for lower undergraduate, Level 10 for a postgraduate or PhD level. If the tool allows, choose the number of results to generate. For example, when generating test questions, you can use the dropdown menu to specify how many the AI should create. These settings can significantly shape the AI’s output, so adjust them to suit your needs.
- Review and edit AI-generated content. Always critically evaluate what the AI produces before using it in your course. The AI can sometimes generate incorrect information, irrelevant material, or phrasing that does not fit your context. Check each AI-generated prompt, question, or image for accuracy, bias, and alignment with your intended learning outcomes. Never assume the AI is completely correct. You should edit the content as needed, for example correct any factual errors, modify wording, or tailor the tone. This step ensures the AI remains a supportive tool under your oversight.
How to use it
While each tool in the Blackboard AI Design Assistant has a slightly different interface, there are some shared steps when you are writing and submitting a prompt. The guidance below focuses on those common steps and features:
- Access the tool: In your Ultra course, choose the appropriate content type (e.g. test, assignment, discussion). Where AI features are available, you will see an option like ‘Auto-generate’.
- Write your prompt: Each tool includes a text box labelled ‘description’ or similar. This is where you tell the AI what you want it to generate. Be specific about the topic, purpose, and any constraints (length, tone, level of detail).
- Adjust general settings: Most tools offer additional options such as complexity level and number of results. Use these to tailor the output to your students and activity. For more detail, please see our guide – Blackboard AI Design Assistant – General settings.
- Add context (optional): Where supported, you can attach course materials to guide the AI’s output. For example, include a reading or handout so that generated questions are based on relevant content.
- Generate and review: Click ‘Generate’ and wait a few moments while the tool processes your request. Review the results carefully for accuracy and relevance before using them.
These steps apply across many of the AI tools in Blackboard, even though the layout may differ slightly depending on what you are creating.
Example Prompt Writing
One of the best ways to understand effective prompt writing is to look at how prompts can vary in quality. Below are examples that show both limited and well-crafted prompts, along with explanations.
Test Question Generation
Limited Prompt: “Give me some questions about psychoanalysis.”
Why it is limited: This prompt lacks specificity. It does not indicate the academic level, the aspects of psychoanalysis to focus on, or what kind of knowledge or thinking the questions should assess. The AI might respond with questions that are too general or miss the key themes of your teaching.
Improved Prompt: “Generate questions suitable for first-year undergraduate psychology students that assess understanding of key principles of Freudian psychoanalysis, including the structure of the psyche (id, ego, superego), defence mechanisms, and the role of the unconscious.”
Why this is better: This prompt defines the subject area, academic level, and specific concepts to focus on. It enables the AI to generate questions that are relevant and pitched at the right level of complexity.

Image Generation
Limited Prompt: “Create an image of a lab.”
Why it is limited: This prompt is extremely broad and lacks detail. The AI will not know what kind of lab, the setting, or the visual style.
Improved Prompt: “Generate an image of a modern genetics laboratory with microscopes and DNA sequencing equipment, in a clean, modern illustration style suitable for use in slides or course materials.”
Why this is better: It specifies the type of lab, the elements to include, and the visual style, leading to a more relevant and usable image.

AI Conversation
Limited Prompt: “Create a role-play related to property law.”
Why it is limited: The prompt does not say who is involved, what the scenario is, or the educational purpose.
Improved Prompt: “Create a role-play conversation designed for second-year law students studying property law. In this scenario, the student takes on the role of a solicitor, and the AI takes on the role of a client involved in a dispute over tenancy rights. The aim is for the student to gather information from the client and explain possible legal options based on UK tenancy law.”
Why this is better: It defines the participants, the topic, and the intended audience, giving the AI clear parameters to work with.

Rubric Generation
Limited Prompt: “Make a rubric for an essay.”
Why it is limited: This is too general. The AI will not know what kind of essay, what criteria to use, or how performance should be evaluated.
Improved Prompt: “Create a rubric for a 2000-word argumentative essay in a third-year philosophy module. Include criteria for clarity of argument, use of sources, critical analysis, and written expression, aligned to UK undergraduate marking bands (e.g. First, 2:1, 2:2, Third).”
Why this is better: It includes essay type, length, academic level, assessment criteria, and institutional context, all of which help the AI generate a useful draft.
These examples highlight the importance of clarity and specificity in prompt writing. The more detail you provide about your teaching context, learning objectives, and content expectations, the more likely the AI will generate something useful and relevant.
Guide last updated on July 4, 2025