FedSubK Feature: ChatGPT & Writing Federal Contract Proposals
I asked ChatGPT to write a blog post of the pros and cons for using itself to write a Federal Government contract proposal. The result may be surprising.
Updated: Oct 2, 2023
I've been thinking a lot about what it's like being a CO in this new era of AI being used for proposal writing. I've been wondering, as a former CO myself for many many years--
❓ What will proposals written with AI look like (or already look like) in terms of technical narratives for specific types of services?
❓ Will evaluation board reviews still be able to easily cull out those proposals that lack nuance or substance, specific to the nature of the work or specific sites / customer locations? Or, will AI add details about work and project sites to a proposal in experience or other areas that a business doesn't actually know or have in an attempt to provide a complete response?
❓ Will proposal responses all start to sound alike for certain repetitive evaluation criteria we see across Government RFPs?
❓ What happens if AI writes a proposal that over-promises on what a business can actually deliver?
❓ What could that mean in terms of mission readiness (or failure)?
❓ Are we sure we are still getting the best contractor for the job when the proposal was not written with the knowledge of the contractor that will perform the work?
❓️ How does the Government know where the offeror's knowledge ends and AI-generated info begins?
💡 I decided I'd go to the proverbial "horse's mouth" and ask ChatGPT to write a blog post of the pros and cons for using itself to write a Federal Government contract proposal. The result may be surprising to some.
------------------------
Title: The Pros and Cons of Using ChatGPT for Writing a Federal Government Contract Proposal
Introduction: In an era dominated by technological advancements, the impact of artificial intelligence (AI) is revolutionizing various industries, including government procurement. One area where AI has shown promising potential is in the creation of written content, such as federal government contract proposals. ChatGPT, a state-of-the-art language model developed by OpenAI, has gained attention for its ability to generate human-like text. In this blog post, we will explore the pros and cons of using ChatGPT for writing a federal government contract proposal.
Pros of Using ChatGPT:
- Time Efficiency: One of the significant advantages of employing ChatGPT for proposal writing is its ability to save time. Traditional proposal writing can be a lengthy and resource-intensive process, involving multiple iterations of drafting, reviewing, and editing. ChatGPT can streamline this process by providing a starting point, generating relevant sections, and suggesting language to use. This time-saving feature can enable government contractors to focus on other critical aspects of their proposal, such as strategy and compliance.
- Enhanced Language and Style: ChatGPT has been trained on vast amounts of text from diverse sources, allowing it to develop a sophisticated understanding of language and style. This expertise can be leveraged to produce well-crafted proposals that are professional, coherent, and persuasive. The model can generate clear and concise language, helping contractors present their ideas in a compelling manner and potentially increase their chances of winning contracts.
- Access to a Wealth of Knowledge: Another advantage of using ChatGPT is its access to a wide range of information and knowledge. The model has been trained on a vast corpus of text, including government documents, industry reports, and legal frameworks. This vast knowledge base can assist in the research and referencing process, ensuring that the proposal is well-informed and aligned with the requirements of the federal government. Contractors can leverage this resource to enhance the accuracy and comprehensiveness of their proposals.
Cons of Using ChatGPT:
- Lack of Industry-Specific Expertise: While ChatGPT has been trained on a vast amount of text, it does not possess real-world experience or domain-specific knowledge. Government contract proposals often require a deep understanding of industry-specific regulations, standards, and technical details. ChatGPT may not be able to provide the nuanced expertise required to address such complex requirements accurately. Human intervention and domain expertise are crucial to ensure the proposal meets the specific needs of the federal government.
- Limited Contextual Understanding: Language models like ChatGPT excel at generating text based on patterns and examples in their training data. However, they may lack contextual understanding, resulting in inaccurate or irrelevant content. Government contract proposals require careful consideration of specific project details, evaluation criteria, and compliance requirements. Relying solely on ChatGPT's suggestions without a human reviewer may lead to inaccuracies, omissions, or misunderstandings.
- Security and Confidentiality Concerns: Federal government contract proposals often contain sensitive and confidential information. Entrusting this information to an AI model raises security concerns. While efforts can be made to anonymize data, the risk of potential breaches or unauthorized access cannot be entirely eliminated. Contractors must carefully evaluate the security measures in place when using AI-powered tools for sensitive content creation.
Conclusion: The use of ChatGPT for writing federal government contract proposals presents both pros and cons. Its time-saving capabilities, enhanced language and style, and access to a wealth of knowledge make it an attractive tool for many contractors. However, the lack of industry-specific expertise, limited contextual understanding, and security concerns highlight the importance of human intervention and domain expertise in the proposal writing process. A thoughtful combination of AI and human oversight can harness the benefits of ChatGPT while ensuring the accuracy, currency, and context of the information provided.
-----------------------------
⚠ My advice...PAY CLOSE ATTENTION TO THE CONS that ChatGPT identified in itself.
Contracting Officers and technical evaluation board members evaluate proposals on depth / breadth of experience, specific knowledge, and context; nuances this tool says it cannot provide. It's fine to use AI for the routine and mundane fluff, but the guts of your proposal--the info that will make or break your success--still must come from you. It's a tool, not the complete solution.
#federalcontracting #proposals #smallbusinesses #rfps #ai
Tags:
View related posts
FedSubK Feature: What is Buying In?
"Buying in". Do you know what that is? Let's illustrate it with a little story...
Once upon a time an agency leader🤴 was looking around at things to make 🌟efficient.🌟 They got the idea that every agency should have the same widgets🔅 their agency had.
The agency leader🤴 called up a widget company👩🔧 and said, "We are interested in your widgets. 🔅What kind of discount can you give us?"
The widget company👩🔧 offers a discount 📉 because they know this agency🤴 not only buys for themselves but may buy for other agencies🫅🤴👸 where a highly trusted widget competitor👨🔧 presently has the work.
The widget company👩🔧 was "buying in" -- offering unrealistic discounts📉 that made the price unrealistically low not only for the current effort but also to influence the purchasing decisions on future buys. Then prices usually up 📈 again over time.
Depending on when "buying in" happens there could also be questions related to compliance with the Competition in Contracting Act (CICA) and possible other violations.
This is why agency announcements that management has made a deal for "$1 a license" and other such management interference is of concern. 🚨 Management plays the numbers game. I'm not saying numbers aren't important, but let's just say... there is a real reason why management typically does not hold contract signature authority. 😬😉
The Government is supposed to keep things fair and do its due diligence. But it's falling for the oldest trick in the book.
Risk, intent, compliance with statutory requirements, misunderstanding of requirements, and comparable market pricing must be evaluated when the Contracting Officer has reason to believe a proposed price is unrealistically low price. But are they?
If a contract isn't in place, there there is still a need to follow appropriate competition rules before a handshake deal. If a contract is already in place, there are things to consider when new discounts appear to be unrealistic including the risk of continued performance, depending on the type of product or service being purchased.
The Government gets a quick win to lock in a low rate, saving some money now. That's called the short game. Government buyers getting blurry-eyed over unbelieveably low prices and don't do the long-term analysis.
But I'll bet you a dollar the company is playing the long game. They are watching and waiting, getting to know your needs and asking loads of questions. "When do you use my widget most?" "Who buys the most widgets?" "When do you typically buy widgets?" And then as fast as they dropped the price, they raise it again on you when you can't afford to make a change -- like at an end of fiscal year. That's how they get locked in and receive perpetual contracts.
BTW...the fairy tale above is a true story. I've had new politicals and new leadership / commanders trot companies into my office saying "Company ABC here says they want to sell us "widgets" at a huge discount compared to what we're paying or others are paying now."
Well...okay then.
As a Contracting Officer, whether I could even begin to entertain that idea depends on several things. It's not an automatic "yes". You could replace "widgets" with just about any product or service and it's probably happened to a Contracting Officer somewhere. Especially as new Administrations come into Government.
The stories in the news that made me think -- "Huh, are they buying in?" are the Axios story "Anthropic wants to sell Claude to the Government for $1". (https://www.axios.com/pro/tech-policy/2025/08/05/ai-anthropic-government-sale-dollar) and FedScoop story "Federal agencies can buy ChatGPT for $1 through GSA deal" (https://fedscoop.com/openai-chatgpt-enterprise-federal-government-gsa-deal-general-services-administration-anthropic/).
My husband (also a retired Contracting Officer) and I look at each other often during the news now and, based on the reported discount or price alone, we know that company is likely "buying in". That's based on our combined 72 years of Fed experience and our Contracting Officer "Spidey sense" from having been around the block a few times. But these deals just the most recent in a series of deals GSA is making with companies since the new Administration came to town. OneGov is the program GSA is, in my former Contracting Officer opinion, using to tout savings under for the press releases. But it may come back later to be a big mistake. I hope I'm wrong.
Program/Project Managers and Contracting Officers AND the competition to these companies...LEARN about it and WATCH for it. It's on the rise.
(And don't get me started on having to argue with new politicals, leadership, and commanders about why I can't terminate a current contract and then turn around and give the same work to another contractor at their unrealistic lower price.🙄😱 That's a topic for another time.)
The practice of "buying in" is becoming more common now. Learn about it and how to spot it.
DoD Reduction In Force (RIF) Guidance
Just when you thought it couldn't get any more confusing, some agencies also have their own RIF guidance separate from the OPM guidance that is what we've heard the most about. DoD is one of those agencies.
A copy of the current DoD RIF guidance, DoD Instruction 1400.25, Volume 351, is found at: https://www.esd.whs.mil/Portals/54/Documents/DD/issuances/140025/140025_V351.PDF?ver=DgEFMmb9dLDV7OV-PLb7VQ%3D%3D
This guide establishes policy, assigns responsibilities, and prescribes procedures for reduction in force (RIF) actions taken under Part 351 of Title 5, Code of Federal Regulations (CFR), as modified by Section 1597(f) of Title 10, United States Code (U.S.C.).
This guidance does not, in full, apply to DoD employees covered by an alternative personnel system (e.g., the Acquisition Demonstration; Science and Technology Reinvention Laboratories; and the Defense Civilian Intelligence Personnel System). Those systems will develop their own policies and procedures for RIF that comply with the law, as approved by the Under Secretary of Defense for Personnel and Readiness (USD(P&R)). This guide also does not apply to Senior Executive Service (SES) positions.
The policy statement in 1.2 states that, "For any RIF of civilians in the competitive and excepted services in the DoD, the determination as to which employees will be separated from employment must be made primarily on the basis of performance."
In accordance with 10 U.S.C. 1597, DoD must report to Congress 45 days prior to implementing an approved RIF.
DoD will comply with 5 CFR 351.402 and 351.403 when establishing competitive areas and competitive levels, respectively. Competitive service employees and excepted service employees are placed on separate retention registers established in accordance with 5 CFR 351.404 and 351.405.
For purposes of DoD RIF, employees are placed in one of two categories:
- employees with a period of assessed performance of less than 12 months, and
- employees with a period of assessed performance of 12 months or more.
An employee’s period of assessed performance for purposes of RIF will be the sum of the months of assessed performance associated with the employee’s performance appraisals within the most recent 4-year period preceding the cutoff date established for the RIF. However, periods of time in a rating cycle for which an employee’s performance was not assessed are not included in the employee’s period of assessed performance.
For example, if an employee receives a rating after serving 10 months of the 12-month cycle, the employee’s period of assessed performance is 10 months for that rating cycle.
For employees absent for military service, periods of time during the rating period may be treated as periods of assessed performance if they meet the requirements of Paragraph 3.3.c.(1) under Paragraph 3.3.b.(2) of the DoD guide.
Retention Factors
Competing employees are listed on a retention register based on--
- Rating of Record. See Section 3.3.c. for rating of record examples based on cutoff dates, military service, time frames for ratings to be used, and ratings from a system other and the Defense Performance Management Program (DPMAP).
- Tenure Group. This follows the definitions found in 5 CFR 351.501(b) for competitive service and 5 CFR 351.502(b) for excepted service.
- Average Score. In general, an employee’s average score for one performance appraisal is derived by dividing the sum of the employee’s performance element ratings by the number of performance elements. The average of the average scores drawn from the two most recent performance appraisals received by the employee, except when the performance appraisal reflects an “unacceptable” rating of record will be reviewed. When the most recent performance appraisal reflects an “unacceptable” rating of record, only that performance appraisal will be considered for purposes of the employee’s average score.
- Veterans’ Preference. This follows the procedures in 5 CFR 351.501(c) with three veterans' preference subgroups:
- AD - 30% or more disabled veteran
- A - eligible for veterans' preference for the purpose of RIF but not for placement in the AD category (i.e., less than 30% disabled veteran determination)
- B - not eligible for veterans' preference for purpose of RIF
- DoD Service Computation Date-Reduction in Force (DoD SCD-RIF). Follows rules of credible service as found in 5 CFR 351.503(a) and (b). DoD does not follow 5 CFR 351.504, which grants additional retention service credit in RIF based on an employee's ratings of record.
Rounds in Reduction in Force (RIF)
Two rounds of RIF will be conducted. Round One, Release from Competitive Level, and Round Two, Assignment Rights, are explained in the document in detail related to types of appointments, order of release from the competitive level, and exceptions that may apply. They are found in sections 3.5 and 3.6, respectively.
Displacement may occur during Round Two. Displacement is the assignment of an employee to a continuing position in a different competitive level that is held by another employee with a lower retention standing (i.e., “bumping” another employee). Displacement may be at the same grade or at a grade up to three grades or grade intervals (or equivalent) below the position of the released employee.
Right of Only One Offer
Employees released from a retention register are only eligible for one offer of assignment (similar to OPM rules), with some exceptions. If the employee accepts and offer, rejects an offer, or fails to reply to an offer in a timely manner, they are not entitled to further offers. However, the DoD Component must make a better offer of assignment to a released employee (i.e., to a position with a higher representative rate) if a position becomes available before, or on, the RIF effective date.
Sample retention registers and scenarios are found in the guide in Appendix 3A. Employees have the right to request a review of retention registers and have representation also be allowed to review the registers, as requested by the employee.
--------------------------------------------------------------------
DISCLAIMER: Info is provided for awareness. I am NOT an HR professional but an HR enthusiast having started in HR and being a Federal supervisor and hiring manager. Before taking any action that changes your status, please seek the advice of an attorney knowledgeable in Federal employment law.
Voluntary Separation Incentive Payment (VSIP)
Voluntary Separation Incentive Payment (VSIP) allows agencies that are downsizing or restructuring to offer employees lump-sum payments up to $25,000 as an incentive to voluntarily separate. The amount received is reduced by Fed and state taxes, social security, and Medicare, as applicable.
The full guide on the program is found at the OPM website https://www.opm.gov/policy-data-oversight/workforce-restructuring/voluntary-separation-incentive-payments/guide.pdf
Eligibility for VSIP requires an employee be employed by an Executive Branch agency for at least three (3) continous years without a time limit and not be--
▶️ a reemployed annuitant;
▶️ otherwise be eligible for disability retirement;
▶️ recipient of a notice of involuntary separation for misconduct or poor performance;
▶️ recipient of any previous VSIP from the Federal Government;
▶️ on a service agreement for which--
➡️ a student loan repayment benefit was paid, or is to be paid, during the 36-months preceding the date of separation;
➡️ a recruitment or relocation incentive was paid, or is to be paid, during the 24-months preceding the date of separation; and
➡️ a retention incentive was paid, or is to be paid, during the 12-months preceding the date of separation.
If you receive a VSIP and later come back to Federal Service within 5 years of the date of the separation on which the VSIP is based, you must repay the entire amount before your first day of reemployment. This includes working under a personal services contract or other direct contract with the Government.
The top 10 questions related to VSIP can be found at https://www.opm.gov/policy-data-oversight/workforce-restructuring/voluntary-early-retirement-authority/top-10-frequently-asked-questions-about-vera-and-vsip.pdf
OPM's page on VSIP is at https://www.opm.gov/policy-data-oversight/workforce-restructuring/voluntary-separation-incentive-payments/
DISCLAIMER: Information is provided for situational awareness. I am not an HR professional but an HR enthusiast having been a Chief of Contracting and Federal supervisor. Please consult with an attorney knowledgeable in Federal employment law before making any decisions that impact your Federal employment status.