AI & Grant Writing, An Update
We’ve been seeing a lot of ads lately promising to teach us how to use ChaptGPT to write grants. When we see those ads, we laugh and laugh and laugh. Those promises are utter and complete hogwash. Two years ago, we published a pair of articles debating the pros and cons of using AI to write grants and a lot has changed since then, so it is time for an update. Despite our initial reservations, over the past 18 months or so, we have just about perfected leveraging AI to expedite grant writing. But note, we say “expedite.” We aren’t using AI to write grants for us. And neither should you. Here’s why…
The Pitfalls of Using AI to Write Grant Proposals
Grant proposals are a cornerstone of funding for nonprofits, researchers, and other organizations. They demand clarity, precision, and a deep understanding of the project's goals, stakeholders, and potential impact. With the rise of artificial intelligence (AI) tools capable of generating text, many are tempted to use these technologies to streamline the grant-writing process. While AI can certainly assist in some aspects, relying on it entirely to write grant proposals introduces significant risks.
Lack of Contextual Understanding and Nuance
AI tools, no matter how advanced, operate within the limits of the data they are trained on. They can generate text that sounds polished and professional, but they lack the ability to understand the nuances of a specific project, organization, or funding opportunity.
For example, grant proposals often require tailoring to the funder's mission, values, and past grants. AI can regurgitate generic phrases like "This project will have a significant impact," but it struggles to weave in the specific language and priorities of the funder. Without this contextual understanding, the proposal may come across as superficial or misaligned with the grant's requirements.
Furthermore, AI does not inherently understand the social, cultural, or ethical implications of a project. A machine-generated proposal might inadvertently use language that is tone-deaf or inappropriate for the target audience, potentially alienating reviewers.
Overreliance on Generic Language
AI tools often produce text that is formulaic and over-reliant on clichés. While grant reviewers are accustomed to standard proposal structures, they also expect originality and specificity. Proposals that feel generic can fail to stand out in a competitive pool.
For instance, AI might generate phrases like "We aim to improve outcomes for underserved communities," but it may not specify how those outcomes will be measured or why the organization is uniquely positioned to achieve them. Reviewers are looking for detailed, actionable plans supported by data, not vague promises. Overuse of generic language can signal a lack of effort or understanding, diminishing the credibility of the proposal.
Inaccuracy and Fabrication of Information
AI tools are not infallible when it comes to accuracy. They can sometimes "hallucinate" facts, creating information that sounds plausible but is entirely fabricated. In grant proposals, where every claim must be backed by evidence, this can be disastrous.
For example, an AI might generate a statement about the effectiveness of a particular intervention but fail to cite a credible source. Worse, it might invent statistics or misinterpret existing data. If reviewers verify the information and discover inaccuracies, the proposal could be disqualified, and the organization's reputation might suffer.
Failure to Capture the Organization’s Voice
Every organization has its own voice, ethos, and mission, which need to come through clearly in a grant proposal. This voice reflects the organization’s values and commitment to its work—something that resonates deeply with funders.
AI-generated text, however, tends to lack personality and emotional resonance. While it can mimic professional writing, it struggles to convey the passion, dedication, and authenticity that human writers bring to the table. A proposal without the organization’s unique voice may feel impersonal, making it harder to build a connection with the reviewers.
…Still with us? Keep reading…
Inability to Strategically Address Reviewer Expectations
Grant proposals are not just about stating the facts—they are about persuasion. Skilled grant writers understand how to anticipate and address the concerns of reviewers, presenting information in a way that strengthens the case for funding. They know how to highlight strengths, address potential weaknesses, and align the proposal with the funder's priorities.
AI, however, lacks the strategic thinking required to craft such a compelling narrative. It cannot anticipate questions like: Why should this project receive funding over others? or What makes this project sustainable in the long term? Without this strategic framing, the proposal may fail to inspire confidence.
Ethical Concerns
Using AI to generate grant proposals raises ethical questions about transparency and accountability. Funders expect proposals to reflect the genuine effort and expertise of the applicant. If an organization submits an AI-written proposal without disclosing it, they risk misleading the funder.
Moreover, reliance on AI could contribute to inequities in the grant application process. Organizations with access to advanced AI tools may gain an unfair advantage, potentially crowding out smaller, grassroots organizations that rely on human effort. This could exacerbate existing disparities in funding distribution.
…Just two more major points to go. You’re doing great!...
Neglect of Collaborative Input
Effective grant proposals are often the result of collaboration among team members, stakeholders, and subject-matter experts. This process ensures that the proposal reflects diverse perspectives and aligns with the organization's broader goals.
AI, on the other hand, operates in isolation. While it can draft text, it cannot replace the rich, iterative process of collaboration. Overreliance on AI might lead to proposals that lack depth and fail to fully capture the collective expertise of the team.
Overlooking the Importance of Editing and Human Oversight
Even if AI is used as a tool to assist in drafting, human oversight is essential. A human writer can refine the proposal, ensuring that it is coherent, engaging, and free of errors. They can also verify that the tone and language are appropriate for the audience.
Without careful editing, AI-generated proposals can include awkward phrasing, redundancy, or even irrelevant content. Rushing to submit an AI-generated draft without thorough review can jeopardize the proposal's chances of success.
Conclusion
While AI can be a useful tool for generating ideas or drafting sections of a grant proposal, it should not replace the expertise, creativity, and strategic thinking of human writers. The risks of using AI—ranging from inaccurate information to a lack of nuance and authenticity—far outweigh the potential benefits when it comes to such a critical task.
Grant proposals are more than just documents; they are a reflection of an organization’s mission, vision, and commitment to making a difference. To ensure success, organizations must prioritize human expertise and collaboration, using AI as a supplementary tool rather than a substitute. By doing so, they can create proposals that are not only well-written but also persuasive and impactful.
…Wasn’t that fun? We asked ChatGPT, “Please write a 750-word article on the pitfalls of using AI to write grant proposals.” It gave us over 1,000 words, everything since “Here’s why…” and our handful of asides.
ChatGPT didn’t lie, this time. But it often does, like the time it told us 27% of a nonprofit’s service population was Hispanic when, in reality, the figure was closer to 5%. You can’t use garbage like that in a grant proposal. ChatGPT specifically is terrible about making up data points and stating them as facts.
We don’t use ChatGPT to expedite grant writing and neither should you. We do know which tools you can use to accurately, reliably, and ethically expedite grant writing and we’re working on a training to share those insights with you. Let us know if you’re interested in the training and share any questions you want to make sure we cover.