Back in school days, finishing a test early wasn’t a flex — it was a red flag. A. Red. Flag. If a student dared to turn in their test early, they were sent back to their desk to check their work. And even when the time was over, many were met with the teacher’s judgy eyes and a commanding voice: Did you check your work?
It felt like punishment at the time. But really, it was a vital lesson.
That skill — tedious as it seemed — was critical then, and it’s a must-have now. With the advent of artificial intelligence systems integrating into our daily operations, we must not assume their work is accurate. We must check our work. This was a hard lesson recently learned by a government contractor.
The Cost of Not Checking
In the Government Accountability Office (GAO) case Oready, LLC, B-423649 et al., September 25, 2025, a contractor submitted a protest that included citations to cases and decisions that simply didn’t exist. The filings appeared to be generated — or at least heavily influenced — by AI. GAO dismissed the protest and issued a stern warning: Do it again, and sanctions are next. The message was unmistakable — accountability still belongs to the human. Let that sink in!
This wasn’t a harmless glitch. GAO harshly criticized the filing as a “gross disregard for our bid protest process.” Bottom line? The contractor didn’t check their work.
Polished ≠ Accurate: How to Audit Your AI Work
In a world where AI can draft, summarize and even argue, the human responsibility remains unchanged: You still have to verify, validate and own the output. AI can assist, but it cannot be held accountable. That’s still our job.
- Verify the sources. If your AI-generated content includes citations, verify each one manually. If you can’t find it, don’t use it.
- Check the logic, not just the grammar. AI can sound polished while being completely wrong. Make sure the logic holds.
- Ask: Does this reflect my thinking? AI can mimic tone, but it can’t replicate your judgment. Own the message.
- Use AI as a collaborator, not a ghostwriter. Let it draft, but you decide what stays. Build a “pause and verify” habit: Before submitting anything AI-assisted, take a moment to ask: “Did I check my work?”
Turns out, those judgy eyes were onto something. Our teachers weren’t just enforcing discipline — they were instilling discernment. In today’s AI-powered landscape, that childhood lesson has become a professional imperative. The tools may evolve, but the principle remains: Check your work. Because when things go wrong, it’s not the AI that gets called into the office. It’s you!
Adeline (Addy) Maissonet is a Procurement Analyst for the Defense Pricing, Contracting, and Acquisition Policy (DPCAP) within the Office of the Secretary of War, U.S. Department of War (DoW). She serves as a senior advisor on contracting policies and procedures and the agency’s representative on the Department’s views on proposed legislation to Congressional members, their staff, and committee staffers. Prior to her current role, Addy served as a Division Chief and Contracting Officer with unlimited warrant authority for the U.S. Army Mission and Installation Contracting Command (MICC) – Fort Eustis, Virginia. Prior to joining the MICC, Addy served as a Branch Head for the Mid-Atlantic Regional Maintenance Center (MARMC), Norfolk, Virginia, with unlimited warrant authority. She also held other procurement positions with the U.S. Navy. Addy holds an MBA in Management and Contracting Level III Certification under the Defense Acquisition Workforce Improvement Act. She is a graduate from Cornell University’s Executive Leadership Certificate Program and Harvard University’s Business Analytics Certificate Program. In her free time, Addy enjoys hiking and overlanding with her family and friends.
Note: The views and opinions expressed in this article are those of the author and do not necessarily reflect the official policy or position of the Department of War.
Leave a Reply
You must be logged in to post a comment.