SKIP AHEAD TO
AI image generation tools allow you to quickly produce high-quality visuals by describing desired images in text prompts. With thoughtful use, these tools can enhance the creation of diagrams, illustrations, and graphics to engage students and enrich lectures and assignments.
Keeping Pace at MIT Sloan: We’re fast, but technology can be faster! We do our best to update our content to match these changes. However, for the most recent software updates and features, please refer directly to the developer’s website. When in doubt, contact us.
Most AI image generators work whereby the user enters text that describes the image they want to produce. After submitting the text, the user views the results and can save the image for their use. In some cases, users can save images in varying sizes and formats.
Editing of images varies by tools. Follow the tool’s instructions. A common method for editing images includes the ability for a user to erase a part of an image they want changed, and then the user updates the description of the image to be more specific about how they want the image to be reflected. For example, a modified prompt could be “The person should be wearing a denim jacket” (instead of a leather jacket). The erased portion will then be updated with a new image derived from the modified text description.
If you want to mimic a style that you’ve seen in other photos, you may find it challenging to replicate it. A tip is to have AI describe the photo that you want to mimic and then use that description as a basis for your image generation prompt. For example, you can upload an image to Google’s Gemini, ask the AI to describe the image, copy that description, modify it, and paste it into your image generator of choice.
Below you’ll find a curated selection of AI image generation tools that have proven valuable in academic settings. Please note that this list is not exhaustive but offers a solid starting point for exploring the capabilities of these technologies.
IMPORTANT: Keep in mind that generative AI tools can produce biased and inaccurate content. Any content generated by AI should be thoroughly verified for accuracy through human review and additional research.
| Tool | Description | Considerations | Pricing | Related Resources |
|---|---|---|---|---|
| Adobe Firefly | Adobe Firefly is a generative AI model designed for high quality image generation and editing. Integrated into the Adobe Creative Cloud, it allows for text-based editing and generation of various media, including images, videos, and 3D content. | Good image generation and a very powerful image editing capability with an intuitive interface that provides multiple output format options. | Free to MIT Faculty, Staff, and Students; Included in MIT’s Adobe cloud license | |
| Microsoft Copilot | Microsoft Copilot is an AI-powered chatbot built on GPT that provides access to DALL-E. It creates high quality images from text prompts. In the enterprise subscription, chat data is not shared with Microsoft or used to train their AI models. | Good image generation with fast generation times. | Free limited tier if you are logged in; Free to MIT Faculty and Staff | |
| ChatGPT | ChatGPT is an AI-powered chatbot built on GPT that provides access to DALL-E. It creates high quality images from text prompts. | Good image generation with fast generation times. | Free limited tier if you are logged in; $20/month paid subscription | |
| Midjourney | A visual AI platform offering diverse image generation and editing features. Producing some of the highest quality realistic images on the market. | High quality output and fast generation times. Currently requires a Discord account to access along with a paid subscription. Quirky interface but has a large supporting community. Note: Midjourney’s Terms of Use state that all images you generate are public property, which means that someone else can potentially re-use your image and there’s nothing you can do about it. | Paid subscription only ($10/month) | |
| Stable Diffusion (Runs on local PC/Mac) | Open Source, text to image generation, wide range of presets, add-on configurations for resolution upscaling, no restrictions on text input. Includes editing capabilities. | One of the most popular AI generation tools. Highly customizable, no prompt restrictions, and you own the copyright to images you create. Runs locally on a PC or Mac with higher end requirements. | Free | |
| Dream Studio (Stability AI) | Text to image and image to image generation based on Stability AI’s Stable Diffusion, wide range of presets, add-on configurations including an editing capability. | Accessible via the web and leverages Stability AI’s most advanced AI models. Offers editing capability, but currently has no available documentation. | Free limited tier; $10 for 1,000 credits (~5,000 images) | |
| Starry.ai | AI-driven tool for artistic image generation, transforming photos into artwork styles. | Accessible via Web or mobile app. Multiple image themes and can create high quality photorealistic images. | Free limited tier; $4.99/week paid subscription |
Explore the wide range of content you can create with these tools such as:
When used thoughtfully, these AI-powered tools can complement human creativity, rather than replacing it. Understanding both the capabilities and limitations of these tools can inform not only your own use of these tools but also your understanding of how students might use them, enabling you to adapt your assignments and assessments accordingly.
As explained in The Legal Issues Presented by Generative AI (2023), “the U.S. Copyright Office recently said that work can be copyrighted in cases where AI assisted with the creation; works wholly created by AI would not be protectable.” Although AI image generation tools offer valuable new creative capabilities, wading into this legal gray area warrants ample caution (Walsh, 2023).
Any content shared with publicly available AI tools should NOT include any non-public data such as sensitive information (e.g., social security numbers, credit card information, or hiring materials) and personally identifiable information to comply with MIT’s Policies & Procedures and the Family Educational Rights and Privacy Act of 1974 (FERPA). To learn more, see Navigating Data Privacy.
Walsh, D. (2023, August 28). The legal issues presented by generative AI. MIT Sloan School of Management. https://mitsloan.mit.edu/ideas-made-to-matter/legal-issues-presented-generative-ai
