Summary
The United States government is currently pursuing various methods of artificial intelligence technology to improve efficiency and integrity within the federal procurement process. Of the areas where the government is seeking to leverage this technology, past performance analysis provides many opportunities for improvement. The government’s systems contain a great deal of data on a contractor’s past performance but no workable avenue by which to assess this data efficiently while conducting responsibility determinations. Further compounding this overwhelming amount of data is the inability of disappointed offerors to challenge affirmative responsibility determinations, creating circumstances where Contracting Officers are conducting insufficient responsibility determinations and contractors cannot hold them accountable. Emerging technology may be able to solve aspects of this issue, but it must be used both responsibly and consistently.
Federal procurement officials are paving the way for the implementation of innovative technology across the government. Examples of this enthusiasm for new approaches can be seen broadly through federal agencies, including the Department of Health and Human Services’ use of artificial intelligence (AI) to consolidate contract vehicles, and the National Aeronautics and Space Administration’s (NASA) robotic programs designed to automate administrative portions of the procurement process. What once took acquisition professionals hours to complete, AI programs can do in minutes, freeing acquisition professionals to expend their efforts on tasks that require critical thought and subjective decision making that AI systems cannot meaningfully replicate. In addition, AI systems have the capability to improve the acquisition process by streamlining and isolating the amount of information that agencies must consider. The constantly evolving procurement field allows for acquisition professionals to adapt leading innovations into usable technology that will benefit both government customers and federal contractors through improving efficiency and data analysis, while also allowing for a potential reduction in corruption. However, AI-driven innovation is not without its risks. For procurement officials to make effective use of this innovation, they must responsibly implement the technology while simultaneously being mindful of its limitations.
Responsibility determinations and past performance evaluations are an essential feature of the contract award process. For responsibility determinations, the Federal Acquisition Regulation (FAR) requires contracting officers (COs) to evaluate several factors that indicate the ability of potential awardees to fulfill the demands of the contract. In conducting the determinations, COs must consider a contractor’s past performance, which requires an investigation into how the contractor has completed work prior to the award of the contract. During the inquiry, the voluminous collection of performance data including different contracts and work that a contractor has taken part in can overwhelm procurement personnel and is thus useless. Yet, by taking advantage of AI programs, government agencies can streamline this complex process and transform contractors’ performance data into useable pieces of information. AI technology provides a promising future for the federal acquisition process, but, to be effective, COs must meaningfully use the AI technology as only part of the process in making informed decisions.
Part II of this Note will discuss the overall recognition for an improvement in how agencies are able to utilize past performance data. It will further explain how agencies use data regarding a contractor’s record of performance to complete a responsibility determination. Then, it will explain why responsibility determinations supported by past performance data are valuable to a procuring agency. Next, it will expand on the discretion that the federal procurement system affords to COs and how reliance on incomplete past performance data can impair COs’ evaluations. Part III of this Note will analyze three initiatives that federal procurement officials have created to address the limitations within past performance evaluations: (1) contractor self-assessment, (2) the Procurement Innovation Lab, and (3) the Determination of Responsibility Bot. Finally, this Note will address how these innovations cannot fully realize their promised utility unless agencies consistently implement the technology, ensure that procurement officials understand how to use the technology and its implications, and provide for meaningful implementation of the technology that is mindful of procurement officials’ discretion.
This part will address how the federal government has sought to implement technology to enhance past performance evaluations in the past and how those efforts need to be consistently updated as technology improves. Next, this part will explain how past performance connects with responsibility determinations and why the determinations are a necessary part of the procurement process. Then, this part will expand on how COs have a high level of discretion when evaluating a contractor’s past performance and how, at times, this discretion has the potential to lead to incomplete or inadequate assessments of contractors.
Across the federal government, officials have recognized the challenges plaguing past performance evaluations. The Office of Federal Procurement Policy (OFPP) has commented on the limitations of the past performance evaluation system and has encouraged continuing efforts in improving its technology. The particular limitations regarding past performance analysis include that untimely evaluations cannot inform COs’ business decisions during a contract award and that COs only conduct past performance evaluations for a small percentage of awards. In response to past performance evaluation issues, the OFPP created targets to assist federal agencies in developing their own better technological approaches and systems to improve past performance evaluations. The OFFP has repeatedly acknowledged that prioritizing efficiency is important because conducting meaningful past performance evaluations allows agencies to ensure that they are selecting products and services that deliver the best value to the federal government.
The FAR requires COs to conduct responsibility determinations. To conduct the determinations, COs must consider seven factors. By evaluating these factors, COs can ensure that the government is obtaining the best value for the government by contracting with offerors that will be able to deliver on the promises made in their proposals. However, COs have broad discretion in determining responsibility, which empowers COs to make decisions more easily, but also restricts disappointed offerors’ ability to hold COs accountable.
The Competition in Contracting Act (CICA) codifies the rule governing responsibility determinations. It states in part that “[n]o purchase or award shall be made unless the contracting officer makes an affirmative determination of responsibility.” Typically, the CO makes this determination right before the award of the contract rather than conducting the determination early in the solicitation process where the CO is potentially considering many offerors. Unlike other evaluation factors that depend on the contractor’s proposal, responsibility determinations are based on the contractor itself. The determinations are independent assessments of a contractor’s ability to perform the contract, and thus they differ from responsiveness determinations or the consideration of past performance as an evaluation factor.
The CO must conclude that the contractor satisfies seven criteria before the contractor will be found responsible. The prospective awardee must (1) have the necessary financial resources for performance or the ability to access these resources; (2) be able to comply with the proposed delivery schedule; (3) have a satisfactory performance record; (4) have a satisfactory record of integrity and business ethics; (5) have the necessary organizational and technical skills or the ability to access them; (6) have the requisite production, construction, and technical equipment and facilities, or access to them; and (7) “be otherwise qualified and eligible to receive an award under applicable laws and regulations.”
In performing this assessment, COs must consider general standards that apply to all contracts, as well as “definitive criteria,” which are special standards that only apply if they are included in the solicitation. COs generally use both standards to assess whether a contractor will be able to successfully perform the requirements of the contract’s performance work statement. Further, COs typically will only evaluate the responsibility of prime contractors and assume that if the COs find that a contractor is responsible, then that prime contractor’s proposed subcontractors will also be responsible. However, COs can also review subcontractor responsibility if they feel the additional review would benefit the determination process. Looking to these seven standards, one can understand the CO’s expectation that they can be confident that awardees will be able to perform. Yet, considering the requisite level of evaluation, one can also imagine how a CO may be constrained by a shortage of time, resources, or both to conduct such an in-depth assessment resulting in a determination that is not necessarily holistic or well-conducted.
The FAR mandates responsibility determinations to support the government’s interest in obtaining “the best value product or service to the customer, while maintaining the public’s trust and fulfilling public policy objectives.” Through their use, the government can ensure that contractors are delivering on the commitments that they make in their bids or offers. By ensuring that contractors are responsible, acquisition professionals are able to protect taxpayer interests which is paramount throughout the procurement process. However, COs can only use these determinations to achieve the government’s goals if the COs have accurate and accessible data to review.
Beyond using responsibility determinations to ensure awardees have the ability to deliver on their promises, responsibility determinations can also be used to reduce corruption. Using responsibility determinations to combat corruption—similar to the aims of suspension and debarment procedures—can result in exclusion decisions. Responsibility determinations differ from exclusion decisions in a few ways, and understanding these differences speaks to the specific utility of responsibility determinations. In terms of timing, the government must first find an offeror responsible before making award, which lets potential contractors demonstrate to the CO throughout the procurement process that they can meet the requirements outlined in the FAR. While responsibility determinations serve to ensure that the government contracts only with responsible contractors, a nonresponsibility determination for one contract award does not prevent the government from determining the contractor responsible for another, separate contract. Therefore, nonresponsibility determinations are different from exclusion determinations resulting from a suspension or a debarment, which has government-wide consequences for the contractor, rather than consequences exclusive to a given contract. Further, responsibility determinations are made by COs, while exclusion decisions are made by agency heads or their designees. However, suspension and debarment information can prove to be valuable in conducting past performance assessments, and procurement officials could potentially use emerging technology similarly to benefit suspension and debarment determinations.
In conduction responsibility determinations, COs have a high of level of discretion. Because of this broad discretion, the Government Accountability Office (GAO) rarely hears protests on this topic. This is despite the fact that responsibility determinations are necessary for contract award and, at times, done in an incomplete manner. Additionally, contractors are not entitled to due process when a CO determines they are not responsible. The GAO will typically only hear protests where the contractor can show evidence that the CO unreasonably failed to consider relevant information or violated a statute. Nonresponsibility determinations can also occur when the CO does not have sufficient information to assess the contractor’s responsibility.
When conducting responsibility determinations, COs consider both information that the contractor submits and information that the agency has gathered on its own terms. Prior to award, the government conducts a survey that depends on the level of experience and familiarity that the government has with that particular contractor.
Further, the FAR requires that COs consider information recorded on SAM.gov and “any other relevant past performance information on the offeror.” The FAR also recommends that COs support their determinations with records and experience of the agency’s contracting office as well as other agency’s offices, information directly from the prospective contractor, commercial information available in the public sector, pre-award survey reports, and publications circulated by entities such as trade associations or other agencies.
COs generally have broad discretion in conducting responsibility determinations. However, Congress limited this discretion with the Clean Contracting Act of 2008, which established the Federal Awardee Performance and Integrity Information System (FAPIIS) and required COs to consult the system when making responsibility determinations. FAPIIS has since been integrated into SAM.gov, and COs can see descriptions of all proceedings that involved federal contracts with findings of fault and nonresponsibility determinations. Therefore, the government has taken steps to empower COs to conduct holistic reviews of a contractor’s ability to perform, yet the government has the potential to take further steps to deepen this assessment through emerging technology.
The Contractor Performance Assessment Reporting System (CPARS) provides information regarding a contractor’s past performance to source selection officials. CPARS is intended to be an objective report that assesses a contractor’s performance and adherence to contract requirements. CPARS provides a promising avenue for technological progress within the federal procurement system. The system seeks to improve transparency regarding past work done by contractors and to put contractor performance data into a readily accessible database. Government officials created the database to aid acquisition professionals in making responsibility determinations and help them better understand if a contractor would be able to deliver on their promises in a more efficient way. Given that past performance is not only used for responsibility determinations but that it is also used as an evaluation factor in many procurements, the government has even more incentive to improve the analysis. CPARS allows COs and other source selection officials to input data regarding past, current, and complete contractor performance. The system then compiles the data into ratings that COs can use to assess contractors. The CPARS ratings and associated data are classified and cannot be accessed through Freedom of Information Act (FOIA) requests. The ratings are only available to relevant acquisition officials and the contractor that is the subject of assessment; notably, other contractors cannot view this information. CPARS ratings specifically generate past performance information (PPI), which acquisition officials utilize to evaluate contractor performance.
Generally, CPARS ratings are disfavored by both acquisition professionals and contractors because reviewing CPARS data is time-consuming, and the data can be inaccurate. Overall, the government and private industry value the idea behind CPARS: COs want to be able to easily access past performance data and contractors want the opportunity to demonstrate their strengths. Yet, the current system is overwhelmed by issues that limit its utility. Because of these issues, and the complexities of the process, COs often give contractors a rating of “satisfactory.” This rating has the potential to fail to account for both outstanding and underperforming. contractors.
As a practical matter, COs and contractors cannot use CPARS data because the system retains a massive amount a data about contractors that is not practical for a CO to manually review in a short amount of time. Additionally, contractors are unable to see competitors’ information in the CPARS database, and COs do not have the time or the resources that would be needed to sift through the amount of data in the system that would make for a holistic responsibility determination. The idea behind using past performance as an indicator of contractor responsibility is logical—the government wants evidence that the contractor has historically been responsible and thus will be able to deliver on its promises. Yet, pragmatically, the government has not maintained the infrastructure necessary to achieve this goal, which has led to COs developing their own forms of past performance assessment through questionnaires, instead of relying on the system. However, there is potential to create a system or process so that COs can digest the high volume of CPARS-generated data in a way that can lead to more comprehensive responsibility determinations.
Given the amount and complexity of data presented to COs regarding past performance, COs may not complete past performance analyses in a meaningful way. On the one hand, considering that past performance is a necessary component of the responsibility determination process, and, because they are performed right before award, responsibility determinations likely cannot take into account all relevant information, and COs do not have the resources or the incentive to do this evaluation thoroughly. On the other hand, waiting until immediately before award to conduct responsibility determinations is not without its benefits. The wealth of data available in CPARS may not prove to be useful in conducting responsibility determinations. COs could attempt to review CPARS data sooner as a part the determinations earlier in the process, but the CO may eliminate potential offerors too early and there may be too much data for this step to be practical. A major tradeoff in waiting until right before award to conduct the responsibility determination, however, comes from a scenario where the CO must make a determination of nonresponsibility after a lengthy procurement. This potential for this circumstance does not incentivize the CO to do a thorough responsibility determination—especially given the amounts of CPARS data—at this point because both the contractor and the government have expended immense resources and starting from the beginning seems to come at a high cost.
Further adding to this challenge is the fact that past performance evaluations inform responsibility determinations, and disappointed offerors largely cannot challenge affirmative responsibility determinations. The GAO defers to an agency’s judgment absent a showing that the CO-made determination is contrary to evidence that the contractor specifically did not meet a material requirement; thus, the courts and the GAO will rarely strike down a CO’s affirmative responsibility determination. However, this passivity could create a situation where COs have minimal incentive to look critically into past performance. This could cause a serious accountability issue where COs could avoid obtaining the best value for the government, and disappointed offerors would have no means by which to hold the government accountable. Therefore, COs may be making decisions based on convenience that ultimately end up harming the government and taxpayers.
This part will explore the three government initiatives to increase the quality of past performance evaluations that largely rely on emerging technology: contractor self-assessment, the CPARS AI initiative, and the Determination of Responsibility Assistant bot (DORA bot). These technological innovations provide a promising future for legitimizing responsibility determinations, but resources, interests, and widespread access have the potential to limit their ability to be either meaningful or effective.
The first initiative addressing the inadequacy within past performance evaluations centers on allowing contractors telling their own story of past performance. On February 25, 2021, Jeffery Koses, a senior procurement executive in the General Service Administration’s (GSA) Office of Acquisition Policy, issued a memorandum to the GSA acquisition workforce to give COs advice for requesting self-assessment from contractors. In this memorandum, Koses likened having contractors submit self-assessments to employees keeping track of their own performance to supplement an annual review. Self-assessment helps the employer, or here the CO, avoid collecting performance data on their own terms. Further, the contractors have already identified their respective strengths. Koses argued that contractor self-assessments provide a dual purpose: (1) they allow for COs to save time by only requiring time for editing contractors’ assessments; and (2) they provide a space for contractors to advocate for themselves. GSA requires that the contractors input their self-assessment in the same format as used for inputting data into CPARS. This inclusion of self-assessment additionally balances out the past performance evaluation process by giving contractors more of a say in the process and the opportunity to defend their performance.
Yet, contractor self-assessment could potentially be problematic because contractors have the opportunity to produce glowing self-assessments by handpicking their best reviews. Contractors offer their own perspective based upon their past performance; however, intuitively, the contractor has a high incentive to use their best reviews and leave out their worst. Although GSA encourages COs to use self-assessments as a starting point for past performance discussions rather than an end, COs have the discretion to disagree with a contractor’s self-assessment. However, it is unclear if this is practical during procurements; while the self-assessments are a theoretically beneficial starting point, they may well allow COs to avoid conducting a meaningful inquiry into the contractor’s self-assessment. This is, again, worsened by the demand put on COs, and it is intuitive that COs might be tempted to use the self-reported CPARS data as an end point and justification for making an award.
GSA is not alone in turning to self-assessment to evaluate contractors. The United States Department of Homeland Security (DHS) is working to implement self-assessment for contractors regarding cybersecurity requirements. In contrast to the Department of Defense’s (DoD) use of third-party assessment for contractors to certify their cybersecurity practices, DHS is planning to permit contractors to self-report. The contractor self-assessment method provides greater accountability than a self-attestation, where the contractor would simply affirm adherence to the requisite cybersecurity standards. However, contractor self-assessment differs from DoD’s third-party assessments in that the government is still placing a degree of trust in the contractor. Furthermore, self-assessment does not include the potential cost burden on contractors that third-party assessments impose. Importantly, DHS has found in its research that contractors were accurately portraying their capabilities, which improves the CO’s efficiency.
Requiring self-assessment for contractors can reduce COs’ workload in terms of determining a contractor’s responsibility. However, self-assessment does not absolve the CO of all obligations to ascertain that the contractor is responsible; the CO would still need to approve the assessment. Further, DHS’s method removes the need for reliance on a third party. While DHS’s self-assessment model is intended for monitoring cybersecurity compliance, it could have broader implications that extend to past performance. COs could use this self-assessment method to streamline the process pre-award. COs need to be held accountable for assessing a contractor’s performance. While self-assessments can be a useful tool in analyzing the contractor, they cannot be the only tool.
DHS has also recognized the limitations present in the realm of past performance analysis. Because DHS, similar to most other federal agencies, uses past performance as both an indicator of contractor’s responsibility and, in some instances, an evaluation factor, it is seeking to better utilize CPARS to gain deeper insight into past performance data. DHS believes strongly in emerging technology’s ability to shape the future for procurement. As a result, it created the Procurement Innovation Lab (PIL) to “help lower entry barriers for non-traditional contractors, shorten the time to award and increase successful outcomes with improved acquisition techniques.” Further, the PIL does not aim for policy changes. Instead, it seeks to use tools already accessible within the FAR, such as CPARS, to make implementing technological innovation more realistic for federal agencies.
As DHS has developed the PIL, it has had far-reaching impacts on procurement policy government-wide. The success of the PIL can be seen from its influence on Congress. In 2021, Congress passed the Promoting Rigorous and Innovative Cost Efficiencies for Federal Procurement and Acquisitions Act of 2021 (PRICE Act) requiring DHS to follow guidance from the PIL regarding procurement. Further, the Act mandates that the Undersecretary of DHS publish an annual report detailing the positive procurement outcomes that have come from the PIL, such as how its innovations and policies have improved small businesses’ ability to engage with federal procurement and how it has enhanced competition. Generally speaking, the Act requires the PIL to report on its best practices at DHS and share how other agencies can take advantage of its ideas and advancements. While the Act centers on supporting small businesses, the impetus placed on the PIL demonstrates the government’s overall confidence in its ability to reform and improve federal procurement practices through a technology-centered approach.
In terms of past performance analysis, the PIL has also recognized the challenges presented by CPARS and agency evaluation of contractor performance data, and it is working to make a meaningful difference in the way that COs look at this data. The PIL is developing means to harness the mountain of data of available to a CO in CPARS during a procurement and create a system that is more valuable to both contractors and acquisition professionals in its totality. In creating this initiative, the PIL recognizes that widespread use of these programs will ultimately result in lower costs for all users, programs that can be shared and utilized across the government, and increased value within the system.
The PIL’s CPARS AI initiative seeks to leverage emerging technology to make the voluminous amounts of past performance data more digestible for COs, which could, in turn, enhance the overall responsibility determination process. The idea is that the technology will be “used to support the knowledge discovery process of a human.” Therefore, the AI initiative is not standing in as a decisionmaker; rather, this technology will support the CO in making a more informed decision. Ideally, once complete, the program will be accessible to all agencies. With this in mind, DHS partnered with the OFPP, Procurement Council on E-Government, GSA’s Integrated Award Environment, Federal Information Technology Subject Matter Experts, and the Chief Acquisition Officers’ Council to broaden the program’s reach by making it more widely accessible. Following DHS’s research that revealed three out of four acquisition professionals believed that past performance evaluation could greatly benefit from automation, DHS sought to realize this goal of increasing efficiency through developing ways that the process could utilize automation.
DHS recognized that, although CPARS has the potential to drastically improve the manner in which procurement officials review past performance data, the program currently contains an overwhelming amount of data that cannot easily be sorted. CPARS has over one million records for over sixty thousand contractors. Further, a contractor may have contracts with several different agencies for numerous products and services resulting in even more complex data for the CO to review. To tackle this issue, DHS asked industry professionals: “How could emerging technology . . . help federal contracting professionals rapidly access relevant records from CPARS for a given source selection; and could it provide adequate insight into the data to aid in the review and evaluation of a vendor’s past performance?” With this challenge in mind, DHS structured a special non-FAR-based procurement under the Commercial Solutions Opening Pilot Program (CSOP) with an expedited timeline that worked with nine contractors to develop AI products.
The CSOP contractors conceptualized the solution to the past performance challenge as similar to the credit report marketplace, where agencies could easily access information about a contractor’s past performance, similar to how private businesses are able to determine credit information commercially. In addition, DHS did not want to develop a system to totally replace CPARS. Instead, its aim was to access and understand the data already present within the system, negating the need to expend resources on an entirely new system. Although both the government and industry recognize the issues within CPARS, they feel that the system still contains useful data and should be improved rather than discarded altogether.
A few key features of the proposed technology include text analysis that can identify inconsistencies between a CO’s notes about a contractor and the overall rating given by the CO. For example, the substantive text of a CO’s notes about a contractor may indicate that the CO believes that its performance is “unsatisfactory,” yet the CO provides an overall rating of “satisfactory.” The CO may provide a misleading rating because of the burden of the manual process, which lends itself to mechanistic comparisons that limit the CO’s ability to think more critical about differences in proposals. The AI technology can flag inconsistencies such as the discrepancies between a CO’s individualized notes and their overall rating of a contractor. The technology can lead to more cohesive or descriptive information about a contractor that can improve a CO’s decision-making process. Notably, the AI technology does not substitute the judgment of the CO; rather, it improves the ability of the CO to make decisions and clarifies a CO’s notes or comments about an individual contractor.
One of the CPARS AI Initiative contract holders, CORMAC, seeks to improve past performance evaluations with the CORMAC Envisioning and Prediction Enhancing System (CREPES). CREPES uses machine learning (ML) and natural language processing (NLP) to sort through past performance data in way that allows for the data to become more digestible for COs. CREPES applies NLP to place an active solicitation in conversation with a prospective contractor’s CPARS data and then produces a score characterizing how well the prospective contractor’s past performance data corresponds to the needs laid out in the solicitation. Essentially, the CREPES technology reviews the solicitation and generates an understanding of what the solicitation is looking for from a potential awardee. Then, CREPES reviews a prospective contractor’s data on its past performance and produces an evaluation based on how the contractor’s data would fit into what the agency wants from an awardee. CORMAC purports that CREPES is able to predict whether a contractor is likely to be successful in performing a contract and to explain the benefits of using a methodology that is not subject to bias or “human intuition.” Finally, the data can be displayed in a way that is user-friendly and visually digestible for COs. In creating an easy-to-understand user interface, systems like CREPES provide promising solutions for many of the issues that acquisition professionals point to within past performance analysis. Once complete, the PIL’s forthcoming Artificial Intelligence for Past Performance Project has the opportunity to provide COs with the potential to make decisions based off the wealth of data in CPARS but in a much more manageable way.
Another area for advancement regarding responsibility determinations comes from the Army’s DORA bot. Rather than replacing the work of COs, emerging technology, like the DORA bot, frees up COs’ time and allows them to work on more meaningful non-administrative tasks. Similar to the idea behind DHS’ PIL program, the Army did not create the DORA bot to replace COs or their ability to make award decisions. Instead, the Army created the DORA bot to assist and support the CO in intelligent decision-making. The DORA bot differs from DHS’s PIL CPARS project in its use of technology to compile data in a more administrative way that completes objective tasks as opposed to using AI-assisted decision making that is more subjective. DHS relies more on ML, whereas the DORA bot uses robotic process automation, meaning that DHS’ technology institutes more of an “understanding” on the part of the technology that leaves the technology to make its own conclusions. In contrast, the DORA bot does not draw conclusions based on the data that it reviews, and, instead, it simply collects and organizes data into a more usable format without drawing any inferences. Further, the technology that DHS is looking to employ is more complex than the technology that the DORA bot uses, but it leaves room for potential error or disagreements between the AI systems and the CO than the DORA bot does.
The DORA bot collects data from the System for Award Management (SAM). Prior to the DORA bot’s creation, this work was done manually by COs in a process that took around an hour. This process has been cut down to nearly five minutes and requires minimal work on the CO’s end, empowering the CO to use this saved time to conduct more thorough examinations into the contractor’s data. Further, the DORA bot minimizes the risk of human error by removing the need for COs to manually input data regarding the specifics of a contractor’s performance record. Overall, the DORA bot supplements the decision-making process that a CO undertakes, rather than replacing it altogether.
To use the DORA bot, a CO first types in a contractor’s Unique Entity ID in an email subject line, which prompts the bot to search SAM.gov and check for registration and active exclusions. Then the DORA bot sends the CO a report summarizing the contractor’s past performance data. Given that the Army issues around 250,000 contract actions each year and must make responsibility determinations at each phase of the procurement, this technology drastically cuts down on the time that a CO must spend entering data. However, COs are still required to look through the data to make their decisions. As such, the Army’s focus in leveraging emerging technology is directed more on administrative efficiency rather than altering the acquisition processes themselves. Moving forward, the Army is looking to expand this technology’s use across DoD.
The promise provided by the important work being done in DHS’s PIL and with the Army’s DORA bot creates an exciting outlook for the future. Emerging technology can streamline the federal procurement process and enhance the end user experience in many ways. However, to realize this future of innovation, the technology must be implemented consistently, responsibly, and meaningfully.
First, to provide realistic utility for acquisition professionals, the government must consistently implement these programs. Albert Sanchez-Graells, an expert in economic law and procurement, writes that “the allure of the potential benefits of deploying digital technologies generates ‘policy irresistibility’ that can capture decision-making by policymakers overly exposed to the promise of technological fixes to recalcitrant governance challenges.” This idea captures the notion that government officials may place too much faith in or maintain overly high expectations with regard to technological innovation. This overenthusiasm places too much emphasis on solving issues with technology that is not necessarily mindful of its complexities or critical of its implications. Further, these tech-based solutions may provide promise for streamlining state and local procurement, but adoption by these other entities may be limited by the fast pace of technological advancements. Thus, agencies, and the government more broadly, should be mindful of the implications of investing in measures that may not adequately fulfill the government’s objectives or that may become quickly outdated.
Second, the government should consider implementing guidance on responsible use for these programs. Similar to the issues with the current CPARS, developing the technology but not providing consistent maintenance does not responsibly adhere to the goals of procurement, which are to obtain the best value for the government through maintaining full and open competition. Reliance on outdated and obsolete technology can limit full and open competition by denying contractors a fair opportunity to compete because the COs are not reviewing the most accurate data in making their decisions. Further, these technological approaches must be used responsibly, meaning that COs should maintain their awareness of how to use this technology and to what extent to rely on it. The government should take steps to ensure that COs understand the use of the new technology and the obligations that accompany it.
Third and finally, these programs must be used in a meaningful way where COs give weight to the data and continue to exercise independent decision-making. COs must critically consider the data available through these programs rather than using them to confirm an inclination to select a certain offeror because the government did not develop this technology to be a replacement for a CO’s judgment. Instead, the government sought its creation to give COs more time to consider the information that they will now have access to instead of spending their time gathering the data.
The above-discussed advancements in procurement technology provide the opportunity for increased efficiency and accuracy within the federal procurement system. Yet, these tech-based answers may not sufficiently address the issues of COs not being held accountable for making inadequate responsibility determinations. This technology does provide for the opportunity for COs to be exposed to more data in an easier-to-understand way, thus enabling them to make determinations based on sound data rather than time constraints or bias.
However, this technology may not necessarily solve the issue of responsibility determinations being insulated from protests. Technology helps to ameliorate the trouble with COs not making complete decisions, but this does not necessarily mean that a CO will meaningfully analyze improved CPARS data. A potential measure for increasing government accountability could come from making certain elements of CPARS data public similar to how the public can access data stored in FAPIIS. In a hypothetical scenario where contractors—likely disappointed offerors—could access CPARS data about their competitors through an interface similar to that of the CREPES product, disappointed offerors may be more likely to protest contract awards. This could provide a more evidence-based method for disappointed offerors to challenge affirmative responsibility determinations as opposed to the current complexities of challenging these determinations. However, government officials could remove the need for contractor-based accountability if COs do meaningfully review CPARS reports and holistically consider them.
The notion of a public-facing CPARS, however, is not possible without emerging technology like the DORA bot and the PIL’s Artificial Intelligence for Past Performance Project. These advancements in technology allow for the potential for even greater accountability within the federal procurement system. By conducting meaningful responsibility determinations through evaluating thorough past performance reports and knowing that disappointed offerors will have access to these reports, COs would likely further be motivated to obtain the best value for the government. Practically, there would potentially be a high degree of concern about permitting contractors to look at their competitors’ past performance data because it could lead to frivolous protests or undermine the authority and deference that acquisition professionals are entitled to. Yet, this open access could be implemented in a way that provides for COs to make meaningful awards based on informed decisions about contractors past performance. However, emerging technology does provide a mechanism by which COs can meaningfully consider past performance data, which has the potential to provide improvements to the federal procurement process.
Emerging technology provides a promising path forward for innovation in government contracting. The many initiatives sweeping across the federal government, and the government’s overall enthusiasm in implementing this technology, evidence enthusiasm for using emerging technology, which can be seen through the passing of the PRICE Act and other policies that encourage acquisition professionals to imagine how technological programs can increase efficiency and reduce corruption within procurement. By increasing the ease of access to past performance data, COs may have more time to consider collateral requirements and take determinations more seriously. This progression provides a mechanism to hold COs more accountable and document award decisions more clearly.
COs can maintain the status quo and retain broad discretion in conducting responsibility determinations and have minimal incentive to do comprehensive evaluations. Yet, failing to adapt to technological advances limits the potential for government contracting to provide the best value for the government. Overall, the government can change the landscape of federal procurement for the better through embracing and exploring the promise of emerging technology. However, this evolution must be done in a way that is meaningful and serves the bests interests of the government and ultimately all taxpayers.
