Don't Let AI Sink Your Grant: How to Use It The Right Way, The Federal Perspective

This article is part of a two-part series on AI and federal grants. Part 1 examines AI use from the federal grantor perspective; Part 2 will explore it from the grantee perspective. 

By: Jason Mistlebauer, Director of Grants and Policy 

April 16, 2026

Federal grantor agencies are increasingly using artificial intelligence to review grant applications, budget narratives, compliance requirements, and reporting submissions. That means applicants who use AI to write grants need to understand how to use it effectively and how federal reviewers are using it to evaluate submissions. 

I felt compelled to write this after attending the REI Systems,  Trachtenberg School of Public Policy & Public Administration, and National Grants Management Association 2026 Annual Grants Management Survey Results Webinar.  

If you work in grants, you have probably had one or more of these conversations by now: can I, should I, or how can I use AI to support my pre-award grants management activities? To answer this question, it’s important to consider the role of AI from the Federal grantor's perspective.  

Here is what that conversation misses: the federal grantor agency at the other end of this equation is asking the same question about its own processes, and in many cases, it has already answered it. They are using AI! Obviously, not to write grants, but to read and evaluate them, to check whether what you are claiming in your budget narrative holds up, whether your costs are allowable, and whether the work you are describing matches what the Notice of Funding Opportunity (NOFO) asked for in the first place. 

I consider myself someone who keeps up with the federal and general grant landscape, including early federal adoption of AI, when individual agencies tracked their business case usage in spreadsheets and posted them on their websites. Still, I was blown away by the conversation in this webinar. It changed things for me, and while Witt O’Brien’s believes in the authenticity of grant application writing without using AI, I think it is unrealistic for everyone in the field to adhere to this philosophy, so if you are going to use AI to write federal grant applications, you need to understand what you are writing for. 

The federal agency is not just reading your application with AI; it is checking it. 

During the webinar, we heard from two Federal grantor agencies that are actively utilizing AI in their grant evaluation and compliance processes. The US Department of the Interior (DOI) uses artificial intelligence to:  

•  Review budget narratives for cost allowability under 2 CFR 200 (Uniform Guidance).
  
  
• Conduct reverse site visits, analyzing documentation submitted by grantees rather than sending staff out in person.
  
  
• Audit compliance with federal administrative requirements
  
  
• Check whether recipients are delivering on the commitments they made in their original applications. 

None of that is theoretical; it is operational. It has a direct implication for anyone writing a federal grant application today: your narrative will be evaluated against specific criteria coded into a system. Vague language and generic program descriptions will not survive that process. If your application reads as if it could have been written about any program anywhere, the federal system and the human reviewer who follows it will notice. 

What goes wrong when applicants lean too hard on AI?

Federal grant reviewers are not debating whether applicants can use AI—they are calling out the downstream consequences when it is used carelessly. Across agencies, reviewers are seeing a pattern: AI-generated content that sounds polished on the surface but fails basic standards of accuracy, accountability, and alignment with program requirements. These failures are not minor editorial issues; they directly affect eligibility, compliance, and an applicant’s credibility with the federal government. 

The problems that federal reviewers are seeing are not subtle. Representatives from the US Department of Agriculture (USDA) shared examples, including AI hallucinations in grant applications and progress reports, fabricated citations, invented research findings, and sources that do not exist in scientific and research-based programs, which cross your entity’s integrity. One, some, or all of these can disqualify an application submission entirely. 

DOI’s concern was slightly different but equally serious. DOI has been receiving progress reports, some of which are clearly AI-generated, that do not match the requirements of the grant project or the original application. When that happens, DOI sends them back. If the applicant continues to submit work that does not align with what they promised, the project and the recipient will be suspended. In the most serious cases, the applicants were terminated. 

Nobody on the webinar panel said that applicants cannot use AI. What they said, clearly and more than once, is that the applicant is responsible for everything submitted under their name. The AI is not. If something is wrong and you do not catch it, that is your problem. 

Managing grants efficiently, without compromising compliance and integrity, can be a challenging task. If your organization is navigating the complexities of grant management, we can help you enhance oversight, streamline processes, ensure outcomes and reduce the risks of waste, fraud, and abuse. Reach out today to learn how our expertise in grants management can ensure your programs meet their goals, stay compliant, and make the best use of taxpayer dollars. 

Authored by: