Using AI Tools Responsibly in Grant Writing
Artificial intelligence tools are becoming part of the academic research workflow. Knowing how to use them responsibly is essential—especially for grant writers. This guide breaks down practical rules for working with AI, based on a recent article about leveraging large language models (LLMs) in grant applications. It also reflects the latest NIH guidance on AI use in proposals.
Rule 1: Always Check the Funder’s Rules
Every funding agency has its own policies regarding AI use. Some permit it with disclosure, others restrict it outright. Before using tools like ChatGPT, review the funder’s official guidance to avoid surprises.
Rule 2: Protect Your Ideas
AI models often save inputs to improve future versions, which risks exposing your unpublished data or unique concepts. Avoid pasting sensitive information such as budget details or specific aims into public chatbots. Instead, opt for secure AI tools offered by your institution or offline models like PrivateGPT or LM Studio. These protect privacy but may have limited capabilities.
Rule 3: Don’t Let AI Write the First Draft
Begin with your own ideas and wording. AI can help polish later drafts, but it shouldn’t create content you haven’t developed yourself. Reviewers can detect generic or impersonal language that doesn’t match your voice.
Rule 4: Be Specific in Your Prompts
Clear, detailed prompts produce better AI feedback. Include the grant type and the section you want help with. Ask for tone, clarity, or grammar checks rather than a full rewrite. Smaller text chunks get more focused responses.
Example prompt: “I’m writing a K99 NIH grant as a postdoc. Here’s a section on my four training goals. Can you help improve clarity and ensure it aligns with the NIH review criteria for the Career Development Plan? Please focus on structure and tone.”
Rule 5: Double-Check Everything
AI sometimes fabricates sources, citations, or data. Always verify every fact and reference it generates. Use AI brainstorming only as a starting point—fact-check before incorporating anything into your proposal.
Rule 6: Use AI as a Helper, Not a Copy-Paste Machine
AI-generated text may contain plagiarism or bias. Treat AI as a brainstorming partner or editor, not a substitute for your own writing. Your expertise and voice should remain front and center.
Rule 7: Learn From the Process
Use AI suggestions as a way to improve your writing skills. Ask yourself why a change works—is it clearer, more concise, or stronger? This back-and-forth sharpens your instincts and helps you become a better grant writer.
Rule 8: Try AI to Brainstorm Visual Ideas
Image-generating AI can help sketch figure concepts or improve visuals. Just confirm that AI-generated images meet funder rules and are scientifically accurate.
Example tools: DALL·E 3 (via ChatGPT Pro), Midjourney (Discord), Adobe Firefly.
Example prompt: “Create a simple conceptual figure showing a timeline of clinical trial phases, including milestones like IRB approval and patient recruitment.”
Rule 9: Human Feedback Still Matters
AI can’t replace human insight. Colleagues and mentors catch issues AI misses, such as scientific logic, cultural tone, or strategic alignment. Combine AI support with human review for the best results.
Rule 10: Play Around and Experiment
Try different AI tools and prompt styles to find what works best. Test prompts on content you know well to understand each model’s strengths and weaknesses.
Bonus: Useful AI Prompt Examples for Grant Writing
- As an experienced NIH grant reviewer, evaluate the following Specific Aims section for clarity, feasibility, and alignment. Highlight weaknesses or ambiguities.
- You are a federal grant coach with 15 years in translational research. Rephrase the Innovation section to emphasize what’s novel and the research gap it fills.
- You are an NIH program officer preparing talking points. Write a 3-sentence summary explaining the project’s value and urgency for a general scientific audience.
- List 3 ways I could visualize the data in this paragraph to show [insert goal].
- Act as a plain language editor. Rewrite this paragraph for a non-specialist reviewer while keeping scientific accuracy.
- You are a professional grant writing expert. Summarize the full draft (Specific Aims, Significance, Innovation, Approach) in a concise paragraph that ties the proposal together.
- Does this paragraph reflect the mission of [insert organization]? If not, suggest edits.
Using AI responsibly in grant writing means combining technology with your own expertise and judgment. For those wanting to deepen their AI skills in writing and other areas, check out Complete AI Training for courses and resources.
Your membership also unlocks:
