ABSTRACT

The Trusted Workforce 2.0 (TW2.0) initiative revamping the personnel security program made strides to implement the process of continuous evaluation (CE). CE isn't a new concept, though, but why did it take so long to implement? I argue TW2.0 built on previous momentum of evolutionary change to capitalize on a moment of punctuated equilibrium while utilizing best practices of designing for new technology. This paper will help to explain why rioters at the US Capitol with security clearances may stay in positions of authority, and it will also raise questions of what policies are needed for clearance reform in the future.

In January 2021, hundreds of individuals rioted at the US Capitol to protest the certification of President-elect Joe Biden.1 The mob falsely asserted that current President Donald Trump had actually won reelection, and rioters took action on current President Trump's rhetoric to march on the capital to protest what they considered a rigged and stolen vote.2 Protesters ignored security features like police guards, barricades, and walls, entering the Capitol building where they broke windows, destroyed property, assaulted law enforcement officers,3 and all the while snapped and posted photos on social media pages.4 By the time the tumult concluded, there were over sixty injured Capitol police officers, five people dead including a Capitol police officer,5 and over 30 million US dollars in damage to the area.6

It is of particular note that some of the rioters even had security clearances, which assumed a greater level of civic responsibility and put them in special positions to access national security information. Reportedly, there was a DEA agent from Los Angeles,7 a retired FBI agent with a top secret security clearance that posted pictures from a Capitol balcony before being charged with crimes connected to the events,8 an Army reservist/naval contractor,9 and a political appointee for the Department of State.10 The potential for clearance holders to have participated at the event was so significant that Congress even initiated an investigation to identify those present with a clearance.11 From a clearance standpoint, the activities of that day could be coded under adjudications criteria addressing loyalty to the United States where someone has “knowingly and willfully engaged in acts or activities designed to overthrow the US Government by force” or otherwise sympathized with those who are attempting to commit those acts.12

The insurrection involving cleared personnel is particularly interesting in context of recent changes in the personnel security program (PSP) and continuous evaluation (CE) (changed to be called continuous vetting in 201913 but referred to as CE for this article) that would have issued clearances to those individuals. Despite decades of status quo, the PSP is actually moving toward organizational transformation. For decades, clearance holders were subjected to infrequent reinvestigations to periodically reevaluate their fitness to hold national security and public trust positions. These investigations were also costly and contributed to an overall backlog in granting clearances, and involvement at an insurrection could have been possibly overlooked for as long as fifteen years. As of 2020, however, there are over 300,000 employees across twenty-six agencies14 enrolled into the CE systems designed to constantly monitor current clearance holders to immediately identify problematic behaviors like being involved in a violent and deadly insurrection. This automation has been credited with an increase in visibility for potentially problematic activities as well as a reduction in backlog in the number of pending investigations.

However, although recent reports of success are positive, ideas for the CE program are not new, and talks of implementation have long-standing precedent. This background thus leads the driving questions of this article: If changing the reinvestigation process was so important, why did it take so long to implement CE? Further, why is this change important to document?

Using this research in information technology and systems, and organizational change, I argue that a review of government documents illustrates a range of social and technical areas of resistance characteristic of organizational inertia, and to overcome this inertia required attention to designing for technological change, which happened for the PSP through the consolidation of power from one agency to another. I argue that those in charge of the PSP were ultimately able to build on previous incremental movements of evolutionary change to capitalize on a moment of punctuated equilibrium while utilizing best practices of designing for new technology. In the backdrop of events such as the January insurrection, this is important not only for legislators interested in implementing technology but also for scholars and citizens thinking about information gathering policies of the future such as those using artificial intelligence (AI) because that is the plan for security clearance procedures to come.15

To come to this conclusion, this paper will be broken into four steps. First, it will offer an explanation of the background investigation process in general. It will then explain organizational inertia and organizational transformation that helps contextualize the documents reviewed for the study. Third, the paper will then present the results of the document reviews, before fourth, discussing how those documents represent working through organizational inertia and organizational transformation.

Background on Background Investigations

First then, it is beneficial to understand the PSP and CE's role in it. CE was an integral part of former President Donald Trump's Trusted Workforce 2.0 (TW2.0) initiative, a technology-heavy plan introduced during his presidency to revamp the information-gathering system of the PSP, and especially the background investigation (BI) process. BIs are instruments key to getting a favorable adjudication for a security clearance that one needs if they hold a sensitive position. The intensity of the BI corresponds to the information classification levels, with the intensity of the investigation on a sliding scale corresponding to the level of classification. The higher the clearance, the more intense the investigation will most likely be. For instance, a top secret clearance investigation to access things like top secret information has lengthy research and interview requirements going back approximately ten years, whereas a clearance for a confidential position may only span five years and require less and more automated depth.16

One important goal of the TW2.0 program and CE was to address what is especially illustrated by those cleared individuals present at the January insurrection: the ongoing fitness of a cleared employee to have continued access to sensitive spaces, equipment, and information. Because it is not enough to give someone a clearance and never revisit that person again, built into the BI process for decades was the step of reinvestigation where a person is subjected to periodic reinvestigation that checks in on a cleared employee's behavior since their last investigation.

One major flaw to this method, though was that it left gaps in coverage. As noted above, previously, those that had a top secret clearance would need a periodic reinvestigation every five years, those with a secret clearance would need one every ten years, and, and too, those with a confidential designation would need one every fifteen years.17 That means after the initial clearance, one would not have a regularly scheduled investigation again for that length of duration. Checks could be triggered by self-reporting or if management was notified in some other way (such as with the DEA agent mentioned above who came to the DEA's attention in the first place “after he sent photographs from his phone to a group chat of fellow agents”18). This means, for instance, if the capital insurrection happened just a few years ago (and still today not everyone is enrolled in CE), clearance holders potentially could be arrested at an event or detained for involvement without triggering an investigation for up to fifteen years. During that time, they could remain in their jobs and could pose a potential threat, for instance, based on the loyalty issue as described above. This would be problematic especially due to access to specialized weapons as in the case of the Army reservist and naval contractor mentioned previously who reportedly had access to a variety of military munitions.19

In part, due to the delays caused by reinvestigations (in addition to other issues like contractor resource problems20 beyond the scope of the paper), in the last fifteen years or so, the PSP has been so problematic that the US government's Congressional “watchdog” agency, the US Government Accountability Office (GAO), has put it on their list of high-risk programs several times.21 The PSP was on GAO's high-risk list first in 2005 “due to a massive backlog of applications and insufficient quality standards”22 and then again in January 2018, when it went back on the high-risk list again23 for processing delays, problems with quality, and IT security.

After this last inclusion onto the high-risk list in 2018, cue the TW2.0 and one of its core programs, CE designed to reduce the backlog of reinvestigations, lower costs, and to eliminate the gaps in breaks in those reinvestigations. One premise of CE was that if the government could automate portions of the reinvestigation step, then technology could be utilized to constantly monitor those with clearances with technology. This would not only reduce the gaps between investigations, but it would also reduce the costs of some of the more manual and lengthy reinvestigation steps, thereby freeing up manpower and reducing the time it took to complete other investigations, in turn also reducing the backlog. Checks would thus be automated and ongoing,24 more frequent,25 could occur at any time,26 fill in gaps between when an issue may go unreported or unknown,27 and would be more agile to react if an issue was identified.28 Those arrested at the Capitol would thus more readily come to the attention of clearance administrators.

By 2020, the CE element of the TW2.0 had been credited especially with reducing a large backlog in clearances,29 which would have increased the timeliness in cases in the PSP in general. With the help of continuous automation, Bill Evanina, the Director of the National Counterintelligence and Security Center said that the agency in charge of the PSP, the Defense Counterintelligence and Security Agency (DCSA), had exceeded expectations, and the agency was ahead of its December 2019 goal to have only 200,000 clearance cases pending, down from 725,000 cases from April 2018.

Important to note, though, that despite that it has been called the “most important element” in President Donald Trump's TW2.0 program,30 CE has been discussed as a solution31 for clearance overhaul for at least the past fifteen years. After 9/11 in response to increased concern about future terror incidents, the US government increased the amount of classified information and spaces, and thus the number of positions that needed a BI went up as well, creating a backlog.32 It was after the 2005 addition inclusion to the GAO's high-risk list that more talk of technological solutions was called for, and CE began to pick up traction. During this time, there were already talks about the existing, but smaller scale automated continuous evaluation system (ACES) developed over the duration of the past twenty years.33 But again, back to the primary research questions, if changing the reinvestigation process was so important, why did it take so long to implement CE?

Organizational Transformation, Inertia, and Strategies for Implementing Change

Preparing to answer this question requires a discussion of three more points: (1) a discussion of organizational and technological transformation; (2) the social and technical nature of organizational inertia in resistance to organizational and/or technological change; and (3) strategies for successfully implementing new technologies.

Organizational and Technological Transformation

Starting with organizational transformation, to make change within an institution, changes have to be made in systems and processes that alter the way people work. Changes can be rapid as in the notion of punctuated equilibrium where change is seen as sudden and systemic, or they can be slower in the case of theories of evolutionism that see transformation as incremental, continuous, and slow.34

Technology can play a part in the transformation of an organization. Technologies help limit and constrain what processes an organization does, and sometimes the technologies that an organization adopts change the fabric of an organization itself, altering the ways people work. A movement of technological, organizational change is illustrated with research on how governments began to use more technology for their public services around the turn of the twenty-first century. In the early 1990s, more and more governmental processes became automated, and scholars noted how “e-government” initiatives redesigned the ways employees worked and how governments worked with the public through processes like automation.35

Resistance to Change Is Social and Technological

Instigating organizational change through technology does not just concern technology, however. Implementing technological change is also a social matter. Second to be discussed then, is that organizations should pay attention to not just their technical needs but also the human and social context where technologies will be used. This is particularly seen through various categories of “organizational inertia.”36 Organizational inertia is what Besson and Rowe37 call the “stickiness” of an organization, and it is one thing that slows down change. Organizational inertia occurs because to implement organizational change, a process must become human/social routine. A routine formalizes practices, though, and the routinized, formalized way of doing things creates inertia because it causes rigidity. Rigid processes are harder to adjust. Processes then get in a cycle, and to change, an organization has to overcome existing routines, but the replacement routines then also reinscribe a resistance to change. Any type of change then has to overcome various forms of resistance. To illustrate, if an agency makes everyone use a new computer program, they will have to overcome previous employee habits and ways of working. These new systems then become the new ways of working that will also eventually have to be phased out when future changes occur.

Besson and Rowe38 identify five categories of organizational inertia that can slow down processes of technical change: psychological, socio-technical, economic, political, and socio-cognitive inertias, showing up at the individual, group, organizational, or sector levels.39 Psychological inertia comes from denial and a fear of learning new things, and socio-technical inertia has to do with dependencies on existing systems and may cause immobility to change due to time needed for development and organizational proclivities for consistency. Economic inertia fights against existing budgets, resources, and the abilities to risk finding and implementing new processes. Political inertia has to do with each stakeholder's existing networks and interests and finally, socio-cognitive inertia relates to group norms seen in individuals, groups, institutions, and societal values.

Important to note, these five points of inertia are both technical and human/social because they illustrate that a technology itself isn't the only factor to consider if one, say, wants to transform an organization. Although more technocentric approaches may assume that employees will adapt to whatever technology is implemented, research also emphasizes that integrating technology is especially complex due to the relationship between technology and the social environment in which it sits, particularly with stakeholders like managers, engineers, or human resources, and their ways of thinking like personal and organizational ideologies, goals, and values.40 Especially since the 1970s,41 researchers have shifted from technocentric arguments of how technologies changed the world to considerations of how society and technology impact each other. Noted during the initial period of e-government, “Technology is fundamentally an organizational and human endeavor linking what is theoretically possible to what actually happens.”42 Technology is thus designed and used when humans and machines work together.

Strategies for Successful Change

Third then, to overcome resistance like organizational inertia, tactics for change need to be both technical and human/social. In their work on designing technological change, Laumann, Nadler, and O'Farrell noted that several factors to consider when designing the implementation of new technology involve leadership; attention to employees (like needs, training, and incentives); consideration of technology; and attention to organizational structures.43 To move toward change, one would need good leadership to make choices and lead the change; consideration of what employees think, know, and work; as well as attention to the particular technologies that were being implemented, and the systems and structures these technologies were being implemented into. Further, organizational transformation can then more readily occur if the systems and ways of doing things are aligned with their environments they serve.44 The new technologies need to work for the people that are using them.

Study and Method

With this background, I move into the argument to address the main research question: if changing the reinvestigation process was so important, why did it take so long to implement CE? I will argue that the PSP faced these areas of resistance. To do this, I used qualitative content analysis of congressional testimony and government documents from June 2013 to July 2015, a time period representing a pivotal stage for the PSP. This was not only a time between when the PSP was taken off the high-risk list the first time in 2011 and before it was placed back on in 2018, but it was also after the 2013 classified information disclosures facilitated by Edward Snowden as well as the workplace shooting involving government contractor Aaron Alexis, both instances that put the PSP under intense scrutiny causing an abundance of US government-held congressional hearings about the PSP as well as motivated the introduction of debated legislation about the PSP.45

To identify relevant texts that is often a first step for content analysis,46 I located twenty-three documents from the US government (congressional hearings, bills, daily editions, and reports/publications) that mentioned the US Congress, BIs, and the security clearance process. The written transcripts of each document were used for each document type. These documents represented a variety of genres for the government, ranging from hearings where topics are debated, to introduced bills, to more formal publications where official finding are summarized and published. Including the variety of genres was useful for not just looking at the official discourse coming from Congress, but also for looking at the more transient debates that were behind the more formalized documents. Further study would be useful for analyzing the potential impact for each type of document. I followed a second step of being “immersed in the data”47 by reading through the materials. Third, I organized categories48 and structured information49 as to look first at what problems Congress had with the PSP, identifying technology as one broad theme. Fourth, I then went through the documents several times to identify, refine, and record content and themes,50 noting which technology was an impediment to CE and then further categorizing those impediments. Fifth, I moved to the next step of drawing conclusions51 as detailed below.

Results

I found at least three broad themes running through the documents. These include debates on the practicalities of manual and automated processes, concerns about technical capacity, and questions on who had authority for the project. The varying viewpoints in all three areas illustrate a divide in both technical and human/social systems further explored in the discussion.

Automated and Manual Systems

First, there were arguments both for automation/against manual processes and for manual work/against automation. I categorized the pro-automation standpoints in five categories: (1) connectivity; (2) quality; (3) risk-awareness; (4) simplification; and (5) timeliness and currency. To explain the first three categories, first, there were arguments that automation can make information easier to access by increasing connectivity needed for the clearances. Electronic systems were argued to be more accessible in general through increased access-validated systems,52 and it would also mean an increase in access to databases of lawfully available information like credit scores, social media, Internet data, and so on.53 There were also calls that the automation be connected to various steps of the BI process and to be shared sharing across agencies.54 Second, automation was also argued as helpful for the quality. If position designations and coverage requirements could be uniform across stakeholders, the quality of the investigation could improve, which could also eliminate waste and offer better services.55 Digital, automated tools would help build in and streamline various quality matters into the steps of the clearance process.56 Third, risk awareness could also be enhanced because automation could keep a better track of individuals involved in the system. A system could monitor personnel in general to see who has an active clearance, and there could be audits of cleared personnel.57 Further, the automated system could also identify high-risk populations in sensitive positions such as those that may have been arrested or are experiencing periods of derogatory credit.58 Fourth, automation could simplify59 the process by making it more streamlined and balanced and requiring less work.60 This could involve a downsized contractor workforce, making the process more balanced in cost, timeliness, and quality,61 and making it more streamlined by eliminating a human element that has to regularly go out and collect records.62 Finally, fifth, automation could also keep information timely and current with responsive systems with real-time information.63 Information quickly retrieved from an automated system would help ensure that up-to-date information was collected on an individual or flag a high-risk individual, which would in turn help with timely processing.64

On the other hand though, there were still arguments for manual processes both for creating a system of automation in general or more specifically for record retrieval. In a general critique, one representative stated, “I am questionable about the Government's ability to automate anything or compute its way out of a paper bag.”65 In more targeted critiques of record review, several pressing issues came up. First, it was mentioned that an in-person visit to an agency might be the only way to get a record.66 There were also concerns that a record obtained directly from an agency may have better information than one obtained through an automated database,67 with some noting the case of workplace shooter Aaron Alexis, important information that could have been obtained only in-person may have been missed.68 I illustrate this by thinking of a local law enforcement agency. The quality of a law enforcement agency's record would be reduced if a data entry clerk only enters basic details of a 300-page report for an online repository, which could result in just a few case highlights versus a full 300-page report one might get if one manually went to an agency to review a file, or even a box of files. The manpower required to either scan or type in whole volumes of documents at the local police level could very well vary due to budget or volume of cases, so not every department has the capability to perfectly automate records. One computer screen may have name, date, and charge, but one file might have a detailed description of the event with witness testimony, photos, court referrals, or even a disposition of what resulted from the charge.

Another complexity was that not all agencies participated in state or nationwide databases. The Chief Security Officer at the US Department of Homeland Security voiced this concern when he stated:

The thing that has troubled me about what, the information we receive from the State and locals and the Federals, is that not all of those agencies contribute to, like, the FBI CJIS system. A person can be arrested in your State, the Palmetto State, and some small police department. They may not be a contributor to the FBI database. So that when we go to do our agency checks and our fingerprint checks, we may not have access to that information of that person's arrest within South Carolina. So, that is a gap.69

Also, even with FBI partnerships, a court disposition might not be connected to a police record,70 so it would be hard to identify the outcome of the arrest.

All in all, though, ideological and practical debates about automation illustrate one complexity that implementing CE would have to overcome. Both the pros of automation and the pros of manual processes reflected two distinct points of view that would need to be worked out when CE was actually implemented.

Technological Capacity and Limits

Debates on the benefits and drawbacks of automation and manual processes weren't the only issue, though. Even though automation has benefits, it was noted that technology can't always meet expectations. The main theme I identified in this area was that even though there are lofty ideas about what automation can do and that other programs have been able to incorporate some forms of automation, a system of CE that meets those requirements is not as simple as just creating software that can meet all Congress' expectations. Just because there are precedents with other databases (among many) like JPAS, Scattered Castles, and CVS that store security clearance information,71 and automated CE with ACES,72 which doesn't mean it is easy to link these databases or that they contain all the data needed to streamline a process. Further, some proprietary information just isn't sharable across agencies despite the 2004 Intelligence Reform and Terrorism Prevention Act advocating sharing. As previously stated, too, there are unlinked law enforcement databases that are in the United States73 and even databases that share sometimes contain the minimum amount of information with incomplete details.74 Even databases that are available must be validated to make sure information provided is of good quality.75 All this work to link and streamline process is also difficult logistically, and work is often underfunded, understaffed, and can sometimes be halted due to budget deficiencies.76 The idea of a massive, interconnected database has some concerned about the very lofty ambitions of the project overall, and some voiced real concerns that the expectations were just not possible.

Overall, for this area, although there was a plethora of discussed databases all doing similar work and pilot tests with preliminary data on how process of CE could work, there was also skepticism that there could be an enterprise-wide database or a database with all the necessary linked databases. At best, too, these systems would be costly both for linkage and maintenance as well as create more costs for information review.

Debates on Legalities and Authority

Finally, there was another thread of discourse voicing concern about policies and standardization of the process for legal and authority-related issues. As evidenced above, there were various other, existing or in process databases that were working toward automation that called into question whose technology was going to be used.77 There was also a general question of who decides what information needs to be incorporated, how to incorporate the information legally, how to verify the veracity of information, and how to regulate the collection, analysis, dissemination, reciprocity, oversight, and costs of the data being collected and utilized for decisions that affect the lives of the clearance holders the systems were expected to regulate.

This area is helpfully seen through the push for an enterprise-wide system. According to House of Representatives hearing, “We must build an enterprise-wide C.E. program that will promote the sharing of trustworthiness, eligibility, and risk data within and across agencies to ensure the information is readily available for analysis and action.”78 This idea kept coming up repeatedly due to not only the call for information sharing in the 2004 Intelligence Reform and Terrorism Prevention Act, but also in the conversations about how various agencies are trying to do their own thing in terms of reforming the PSP and creating redundant technology that can't be used by all parties. Although there was a push for this large enterprise-wide system, there was also a question of how to go about this and how to establish ideas of reciprocity where the results of one investigation using one system of data would meet the expectations of other agencies that may utilize other methods.

Discussion

Taken together, discussions about automation, the system limits of technology, and debates on the authority of CE, all offer explanations as to why there was a delay in implementing CE into the PSP. Implementing this new process of CE wasn't just about thinking of a solution and jamming technology into an organization. It was about debating if automation was the right step, thinking about what technology could actually do, and working out who would be in charge of making and running the process. Through the framework outlined earlier in this article, it can also be explained as organizational inertia in its various forms, be it psychological, socio-technical, economic, political, or socio-cognitive. It is important to note the connections between the results and these concepts because the recommendations for designing for successful technological change require attention to both technical and social issues.

First and second, there were intertwined echoes of both psychological inertia (the denial and a fear of learning new things) and socio-technical inertia (the dependency on the way things have been done to achieve the necessary goals). These facets are especially illustrated by the concerns over automating police records. Psychologically, there were very justified fears of having to work with a new system that socio-technically may not deliver on the necessary case requirements. The present way of doing things involved investigators manually retrieving documents from various record providers. The databases here would be controlled by the agencies through which the records were stored. However, automation would reduce the effort needed to acquire a record by consolidating a record search through the agency's own systems. However, there were also fears that this would also affect the necessary quality goals of the PSP. As stated, some records might not show up in an FBI database if the agency does not cooperate with service, and too, sometimes a court disposition might not be included in the court record. Too, even if there is a large, connected database of contributing points of data like arrest records or court cases, the databases need to be validated to make sure that the quality of document is not reduced on the way from a physical document to a digital, automated document. Thus, integrating CE would have to work through resistance to the legitimate psychological fear of new systems and a socio-technical shift in the way records were gathered and recorded to ensure program goals were being met.

Third, there was economic inertia present (which fights against existing budgets, resources, and risk). Integrating CE was especially economic in nature because revamping systems is time-consuming, and to hire personnel to develop more reliable, connected systems is a cost. Change-makers would have to combat budgets and tap resources. Further, changes in processes would cause at least temporary dips in productivity as the employees learning new technologies work through psychological and socio-technical inertia.

Finally, fourth and fifth, there was political inertia present (which concerns the various stakeholders' existing networks and interests) and socio-cognitive inertia (related to group norms ranging from individual values, group values, institutions, and societies). A political concern was that there was battle to determine the practical logistics of how change would be designed, implemented, and managed. There was confusion as to who was in charge, whose systems would be used, and what legal authorities were granted for various steps of the processes. Friction between agencies and the reluctance to share information would certainly affect the ways that technologies and CE could function, and this is especially apparent after 9/11 when lack of cooperation and lack of interagency information sharing were identified as key contributors to the terrorist events.79 These political frictions could also be seen as socio-cognitive frictions as the various groups of stakeholders bring the values and needs of each of their agencies into consideration of how CE can, should, and should not be implemented.

An Explanation of Success

Now that these documents offer an explanation as to why implementing CE into the PSP was a slow process, it is useful to go back now to the main question again, why did it take so long to implement CE? It was a slow process, yes, but how did those in charge of the clearance process finally move past these issues of inertia to implement CE into the PSP? I argue that bringing CE into the PSP was finally possible when the Suitability and Security Clearance Performance Accountability Council leading the TW2.0 was able to build on the previous momentum of evolutionary change to capitalize on a moment of punctuated equilibrium while utilizing best practices of designing for new technology that focused on both human/social and technological resistance.

To explain more simply, change started out as evolutionary because, as stated, some form of CE had been discussed for decades. The PSP was preparing for change as evidenced by the hearings discussing what needed improvement in the process. However, in October 2019, one major change took effect that could be explained as a point of punctuated equilibrium. The Department of Defense's (DoD) DCSA took over control of the PSP from the National Background Investigation Bureau, which had been doing the clearances since 2005.80 This change was so transformational that the plan was called one of the biggest technological and clearance system overhauls in the past fifty years,81 with CE in particular being “foundational in major policy reform initiatives and TW2.0”82 and “a major technology undertaking, rolling in data across seven categories, and for 29 separate departments.”83

In turn, this newly consolidated home also provided a new environment to create systems and ways of doing things that followed strategies of successful technological change. As discussed previously, Laumann, Nadler, and O'Farrell noted several important factors to consider when designing the implementation of new technology. In particular, to the change in the PSP, the shift provided new leadership, reconsideration of employee needs in general, and attention to the particular technologies used and the organizational structures they would be immersed in.

First, attention to leadership showed up when leaders having led other previous, successful personnel reform initiatives were tasked with leading this change.84 The government also published what has been called a “guidebook” for agencies involved of various steps of the process to help them “reimagine the way they think about establishing trust with members of the federal workforce.”85 Senior official Bill Evanina also mentioned that the success of revamping the PSP was due to “the ”˜almost seamless”™ transition of the OPM's National Background Investigations Bureau into DCSA within the Department of Defense,”86 which would have involved management working together to redistribute work tasks.

Employee needs were also addressed because there was a concentrated effort to change relationships between stakeholders and employees, bringing them into more similar missions and spaces.87 Leading official in the change, Patricia Stokes noted a strategy for the change was to “Communicate, communicate, communicate.”88 Evanina continued that one big and difficult step in the process was to push change in government culture and develop trust between the agencies. Although it would be a “big lift” to do so, it was also a high priority.89 Getting everyone to unite behind one goal was also a factor in the early success, with Evanina adding that the TW2.0 initiative “is the first time ever that the legislative and executive branches are on the same page with regard to clearance reform.”90 Too, it was reported that a critical aspect of the success of the new agency was transferring the OPM employees over to DCSA, and to move over 3,000 employees from OPM to DSCU.91 Further, according to reports, DCSA was able to establish a headquarters in Quantico, VA, as well as consolidate “13,500 employees in 167 field locations conducting BI and adjudicative functions for 105 government agencies.”92

Finally, the consolidation also allowed for reconsiderations of both systems and the technologies needed to meet more unified needs. There was success in centralizing tracking and clearance eligibility and CE for 4.1 million people,93 and the DCSA expects to soon consolidate all the personnel security data from distributed IT databases (although DCSA isn't scheduled for full ownership until 202194). Overall, then, the TW2.0's goals for CE were able to capitalize on a moment of change by building on previous work and focusing attention to both technical and social factors.

Why Is This Important?

To return to the second half of the research question, though, why is this change important to document? There are at least two reasons. First, this conversation is particularly useful for explaining the consequences of the January 2021 insurrection, or more accurately, the potential lack of consequences. The US FBI sets up a specific page in their “Most Wanted” section to help identify a slew of individuals who participated,95 and, in the future, if there are any conversations as to why a rioter with a clearance was not removed from their position despite possible publicity of their acts, the gaps in oversight explained in this paper might offer insights. As noted, millions have a clearance, but now, only 300,000 are enrolled in continuous vetting. Some could potentially fly under the radar if not brought to an administrator's attention.

Further, the gap in this system is also important for looking toward the future. Yes, the current change from a sporadic, human-completed reinvestigation to a more automated, CE program represents a radical change from what has been in place for decades. However, although CE is becoming a reality, there are also visions not just of a TW2.0, but also a TW2.5 or 3.0 program,96 and one that will move more AI into the decision-making processes. As more AI is moved into the BI process, decisions by technology are built into technological systems,97 and there are increased corporate partnerships in data surveillance,98 it is increasingly important to evaluate how informational acquisition is set up, standardized, and managed through policy. Again, standardization leads to rigidity, and rigidity leads to organizational inertia, and future use of AI will become the standardized process difficult to change.

The importance of considering the future of AI in the PSP is especially illustrated by one interesting complexity of the CE process that is also very relevant to the example of the January 2021 insurrection at the US Capitol: social media is not yet a part of the CE process. Among other things, one recommendation given by the GAO was that CE could incorporate social media information, to include other publicly available sources of intel.99 But, to the contrary, there were other reports that “[s]ocial media won't be a component that perhaps observers or members of Congress initially thought it would in the government's efforts to track and monitor clearance holders.”100 Instead, it still has the goal of being used, but it is not currently planned at the level of CE.

This lack of regular social media inclusion could be problematic for national security matters, however, as was illustrated by the insurrection. Social media can identify involvement in destructive behaviors as illustrated by several of the cases of suspect indication through social media platforms.101 Without the inclusion of social media into CE, absent manual notification of activities, current clearance holders could at least in theory fly under the radar of visibility and participate in questionable acts if they are not investigated at intervals by human investigators. For those at the insurrection, this could allow participation in anti-governmental or other problematic activities that could be undetected for an undetermined length of time, especially if CE and continuous vetting are supposed to supplement or replace the reinvestigation.102

Companies are marketing AI and machine learning as a tool to manage social media for BI's, though,103 and especially since 2019, there has been talk of using AI to augment decision-making processes.104 Thinking back at the past and how scholars debated and analyzed the e-government and increasing automation to government practices,105 this conversation not only reiterates the importance of policy for inclusion of new technologies within organizations, but also takes a zoomed out look to more societal questions to ask why and how changes should be implemented. This is especially a concern because as discussed, processes that become routinized are often hard to overcome. Thus, do we want more AI if it can help reach national security goals, and if so, how do we standardize this through policy? What would be the limitations? As one program executive voiced about AI for the PSP, “At the end of the day, a human life is affected.”106 For that individual, they recommended that a human make the final decision regarding who does and who doesn't get a clearance. Future research would be useful in comparing the steps of a process that does use AI to assess when and where it is used and how it can impact final decision-making choices.

Conclusion

For those looking for examples of organizational change in general, or those looking to pinpoint policy moments of transformation within the clearance process, this case is useful for review. Overall, I argue they were able to build on previous evolutionary change to capitalize on a moment of punctuated equilibrium while utilizing best practices of designing for new technology, a change that required attention to both social and technical factors. Although continuously evaluating an individual had been planned for years, the step finally began to be realized, built on the work of decades past but capitalizing on the organizational change to overcome areas of institutional inertia. For the clearance process itself, this conversation gives a window into peculiarities of the clearance process that can allow one to, say, participate in the January Capitol insurrection without losing one's job. For future clearance holders and policy makers, this conversation also calls for future research on how future gaps in the clearance process can and should be addressed, whether with or without AI, because as shown, especially for clearances, once procedures become routinized, it can take years to revise the process.

Acknowledgments

This project has received funding from the European Union's Horizon 2020 research and innovation program under the Marie Sklodowska Curie grant agreement No 707404.

FOOTNOTES

1.

Reuters.

2.

Jacobo.

3.

Reuters.

4.

Andrews.

5.

Evelyn.

6.

Chappell.

7.

Heath.

8.

Polantz.

9.

Marcus.

10.

Shepherd, Hudson, and Hsu.

11.

US Congress, House, Permanent Select Committee on Intelligence.

12.

US Office of Personnel Management, “Introduction of Credentialing, Suitability.”

13.

Ogrysko, “Upcoming Policy Changes Shaping.”

14.

Nyczepir, “As DCSA Surpasses Background Investigation Goal.”

15.

Nyczepir, “AI Enters the Equation in Security Clearance Overhaul.”

16.

Henderson.

17.

US Department of Defense””Defense Security Service.

18.

Heath.

19.

Marcus.

20.

Chabrow.

21.

US Government Accountability Office, “Substantial Efforts Needed.”

22.

US Congress, Senate, Committee on Homeland Security and Governmental Affairs, “Security Clearance Reform.”

23.

US Government Accountability Office, “Substantial Efforts Needed,” 283.

24.

Defense Counterintelligence and Security Agency.

25.

US Government Accountability Office, “Personnel Security Clearances,” 8.

26.

US Government Accountability Office, “Substantial Efforts Needed,” 172.

27.

US Government Accountability Office, “Personnel Security Clearances,” 8.

28.

Kyzer, “Periodic Reinvestigations Are Out.”

29.

Nyczepir, “As DCSA Surpasses Background Investigation Goal.”

30.

Kyzer, “Periodic Reinvestigations Are Out.”

31.

US Government Accountability Office, “GAO's 2005 High-Risk Update.”

32.

US Congress, Senate, Committee on Homeland Security and Governmental Affairs, “Security Clearance Reform.”

33.

Herbig, Zimmerman, and Chandler.

34.

Besson and Rowe, 103”“24, 104.

35.

Laumann, Nadler, and O'Farrell 1”“14, 4.

36.

Besson and Rowe, 104.

37.

Ibid.

38.

Ibid.

39.

Laumann, Nadler, and O'Farrell, 5.

40.

Ibid.

41.

Gorur et al.

42.

Laumann, Nadler, and O'Farrell, 4.

43.

Ibid.

44.

Besson and Rowe, 117.

45.

US Congress, Senate, “S. 113-283 - Enhanced Security Clearance Act of 2014.”

46.

Elo and Kyngäs, 107”“15.

47.

Ibid., 109.

48.

Mayring.

49.

Schreier, 170”“84, 176.

50.

Ibid., 177.

51.

Elo and Kyngäs, 110

52.

US Congress, Senate, Committee on Homeland Security and Governmental Affairs, “The Navy Yard Tragedy.”

53.

US Congress, House, “H.R. 490 - Security Clearance Reform Act of 2015.”

54.

US Congress, Senate, Committee on Homeland Security and Governmental Affairs, “The Navy Yard Tragedy.”

55.

US Congress, Senate, “Safeguarding Our Nation's Secrets.”

56.

US Congress, Senate, Committee on Homeland Security and Governmental Affairs, “The Navy Yard Tragedy,” 118.

57.

US Congress, Senate, “S. 113-283 - Enhanced Security Clearance Act of 2014.”

58.

Ibid.

59.

US Congress, House, Committee on Oversight and Government Reform.

60.

US Congress, Senate, “Safeguarding Our Nation's Secrets.”

61.

Ibid.

62.

US Congress, House, Committee on Oversight and Government Reform.

63.

US Congress, Senate, “S. 113-283 - Enhanced Security Clearance Act of 2014.”

64.

Ibid.

65.

US Congress, House, Committee on Oversight and Government Reform, 61.

66.

US Congress, Senate, Committee on Homeland Security and Governmental Affairs, “The Navy Yard Tragedy,” 31.

67.

US Congress, House, “The Insider Threat to Homeland Security.”

68.

US Congress, Senate, Committee on Homeland Security and Governmental Affairs, “The Navy Yard Tragedy.”

69.

US Congress, House, Subcommittee on Oversight and Management Efficiency of the Committee on Homeland Security.

70.

US Congress, Senate, Committee on Homeland Security and Governmental Affairs, “The Navy Yard Tragedy.”

71.

Ibid., 123.

72.

Herbig, Zimmerman, and Chandler.

73.

US Congress, House, “The Insider Threat to Homeland Security,” 43.

74.

US Government Accountability Office, “Security Clearances: Additional Mechanisms.”

75.

US Congress, Senate, “Safeguarding Our Nation's Secrets,” 90.

76.

US Congress, House, “The Insider Threat to Homeland Security,” 36.

77.

US Congress, Senate, “Safeguarding Our Nation's Secrets,” 55.

78.

US Congress, House, “The Insider Threat to Homeland Security,” 17.

79.

National Commission on Terrorist Attacks.

80.

US Office of Personnel Management, “OPM Prepares for Transfer.”

81.

Boyd, “The Security.”

82.

Defense Security Service, 8.

83.

Kyzer, “Periodic Reinvestigations Are Out.”

84.

Defense Security Service.

85.

Ogrysko, “OPM Details Core Values.”

86.

Nyczepir, “As DCSA Surpasses Background Investigation Goal.”

87.

Defense Security Service.

88.

Ibid., 11.

89.

Boyd, “The Security.”

90.

Ibid.

91.

Kyzer, “OPM Cuts Security Clearance Backlog in Half.”

92.

Hakamaa.

93.

Ibid.

94.

Boyd, “The Defense Counterintelligence.”

95.

US Federal Bureau of Investigation.

96.

Nyczepir, “As DCSA Surpasses Background Investigation Goal.”

97.

Nyczepir, “AI Enters the Equation in Security Clearance Overhaul.”

98.

Heckman.

99.

US Government Accountability Office, “Personnel Security Clearances.”

100.

Ogrysko, “The Future of Continuous Evaluation.”

101.

Andrews.

102.

Defense Security Service.

103.

W., D.

104.

Nyczepir, “AI Enters the Equation in Security Clearance Overhaul.”

105.

Roblek et al., 275”“302.

106.

Nyczepir, “AI Enters the Equation in Security Clearance Overhaul.”

BIBLIOGRAPHY

Andrews, Travis M.
“The Capitol Rioters Kept Posting Incriminating Things on Social Media. Unsurprisingly, They Were Mocked—And Arrested.”
Last modified
January
19
,
2021
. https://www.washingtonpost.com/technology/2021/01/19/capitol-riot-social-media/. Last accessed March 10, 2021.
Besson, Patrick and Frantz Rowe.
“Strategizing Information Systems-Enabled Organizational Transformation: A Transdisciplinary Review and New Directions.”
Journal of Strategic Information Systems
21
(
2012
):
103
24
, 104.
Boyd, Aaron.
“DOD Needs IT Support for Security Clearance Background Investigations Program,”
Last modified
November
20
,
2020
. https://www.nextgov.com/cio-briefing/2020/11/dod-needs-it-support-security-clearance-background-investigations-program/170231/. Accessed March 16, 2021.
Boyd, Aaron.
“The Security Clearance Process Is About to Get Its Biggest Overhaul in 50 Years.”
Last modified
February
28
,
2019
. https://www.nextgov.com/cio-briefing/2019/02/security-clearance-process-about-get-its-biggest-overhaul-50-years/155229/. Accessed March 5, 2021.
Chabrow, Eric.
“What’s Behind OPM’s Ousting of USIS?”
Last modified
September
12
,
2014
. https://www.databreachtoday.com/whats-behind-opms-ousting-usis-a-7315. Accessed March 13, 2021.
Chappell, Bill.
“Architect of the Capitol Outlines $30 Million in Damages from the Pro-Trump Riot.”
Last modified
February
24
,
2021
. https://www.npr.org/sections/insurrection-at-the-capitol/2021/02/24/970977612/architect-of-the-capitol-outlines-30-million-in-damages-from-pro-trump-riot. Accessed May 28, 2021.
Defense Counterintelligence and Security Agency
.
“Continuous Vetting.”
Accessed
July
15
,
2020
. https://www.dcsa.mil/mc/pv/cv/. Accessed July 15, 2020.
Defense Security Service
.
“A Q & A with Patricia Stokes.”
DSS Access
8
, no.
2
(
2019
):
8
. https://www.dcsa.mil/Portals/69/documents/about/err/DSS_ACCESS_v8i2.pdf. Accessed July 15, 2020.
Elo, Satu and Helvi Kyngäs.
“The Qualitative Content Analysis Process.”
Journal of Advanced Nursing
62
, no.
1
(
2007
):
107
15
. doi:.
Evelyn, Kenya.
“Capitol Attack: The Five People Who Died.”
Last modified
January
8
,
2021
. https://www.theguardian.com/us-news/2021/jan/08/capitol-attack-police-officer-five-deaths. Accessed March 16, 2021.
Gorur, Radhika, Mary Hamilton, Christian Lundahl, and Elin Sundstrom Sjodin.
“Politics by Other Means? STS and Research in Education.”
Discourse: Studies in the Cultural Politics of Education
40
, no.
1
(
2019
):
1
15
. doi:.
Hakamaa, Mario.
“History of How the Defense Counterintelligence and Security Agency Came to Be.”
Last modified
October
30
,
2019
. https://www.clearancejobsblog.com/history-of-how-the-defense-counterintelligence-and-security-agency-came-to-be/. Accessed March 16, 2021.
Heath, Brad.
“U.S. Suspends Federal Agent Who Joined Crowd Outside Capitol During Rampage, Lawyer Says.”
Last updated
March
3
,
2021
. https://www.reuters.com/article/us-usa-capitol-agent-idUSKCN2AV2XC. Accessed May 28, 2021.
Heckman, Jory.
“ODNI Previews Updated Counterintelligence Strategy, Trusted Workforce 2.0 Rollout.”
Last Modified
February
5
,
2020
. https://federalnewsnetwork.com/cybersecurity/2020/02/odni-previews-updated-counterintelligence-strategy-trusted-workforce-2-0-rollout/. Accessed May 28, 2021.
Henderson, William H.
“Security Clearance Frequently Asked Questions—How Long Does a Security Clearance Remain in Effect?.”
https://www.clearancejobs.com/security-clearance-faqs. Accessed September 29, 2020.
Herbig, Katherine L., Ray A. Zimmerman, and Callie J. Chandler.
“Evolution of the Automated Continuous Evaluation System (ACES) for Personnel Security.”
Report no. 13-06,
November
2013
. Accessed March 13, 2021. https://apps.dtic.mil/dtic/tr/fulltext/u2/a626819.pdf.
Jacobo, Julia.
“This is What Trump Told His Supporters before Many Stormed Capitol Hill.”
Last modified
January
7
,
2021
. https://abcnews.go.com/Politics/trump-told-supporters-stormed-capitol-hill/story?id=75110558. Accessed May 28, 2021.
Kyzer, Lindy.
“OPM Cuts Security Clearance Backlog in Half, But Processing Delays Spell Trouble for Pentagon.”
Last updated
July
22
,
2019
. https://www.govexec.com/management/2019/07/opm-cuts-security-clearance-backlog-half-processing-delays-spell-trouble-pentagon/158586/. Accessed May 28, 2021.
Kyzer, Lindy.
“Periodic Reinvestigations are Out, Continuous Vetting Is in for Security Clearance Holders.”
Last modified
March
11
,
2020
. https://www.govexec.com/management/2020/03/periodic-reinvestigations-are-out-continuous-vetting-security-clearance-holders/163695/. Accessed May 28, 2021.
Laumann, Edward O., Gerald Nadler, and Brigid O’Farrell.
“Designing for Technological Change: People in The Process.”
In
People and Technology in the Workplace
,
1
14
.
Washington, DC
:
National Academy Press
,
1991
. https://www.nap.edu/catalog/1860/people-and-technology-in-the-workplace. Accessed May 28, 2021.
Marcus, Samantha.
“N.J. Naval Contractor Charged in Capitol Riot after Allegedly Bragging to Informant.”
Last modified
January
18
,
2021
. https://www.nj.com/news/2021/01/nj-naval-contractor-charged-in-capitol-riot-after-allegedly-bragging-to-informant.html. Accessed May 28, 2021.
Mayring, Philipp.
“Qualitative Content Analysis.”
Forum: Qualitative Social Research Sozialforschung
1
, no.
2
(
2000
). http://www.qualitative-research.net/index.php/fqs/ article/view/1089/2385.
National Commission on Terrorist Attacks
.
The 9/11 Commission Report
.
Washington, DC
:
Government Publishing Office
,
2004
.
Nyczepir, Dave.
“AI Enters the Equation in Security Clearance Overhaul,”
last modified
April
15
,
2019
, https://www.fedscoop.com/ai-federal-security-clearance/. Accessed March 13, 2021.
Nyczepir, Dave.
“As DCSA Surpasses Background Investigation Goal, is Trusted Workforce 2.5 Likely?”
Last modified
May
14
,
2020
. https://www.fedscoop.com/dcsa-background-investigations-goal/. Accessed March 12, 2021.
Ogrysko, Nicole.
“The Future of Continuous Evaluation is Just About Here, and It Has a Different Name.”
Last modified
September
6
,
2019
. https://federalnewsnetwork.com/workforce/2019/09/the-future-of-continuous-evaluation-is-just-about-here-and-it-has-a-different-name/. Accessed May 28, 2021.
Ogrysko, Nicole.
“OPM Details Core Values behind Coming Security Clearance Reforms with New Policy Doctrine.”
Last modified
January
12
,
2021
. https://federalnewsnetwork.com/workforce/2021/01/opm-details-core-values-behind-coming-security-clearance- reforms-with-new-policy-doctrine/. Accessed May 28, 2021.
Ogrysko, Nicole.
“Upcoming Policy Changes Shaping DoD’s Newly Renamed Security Clearance Agency.”
Last modified
July
19
,
2019
. https://federalnewsnetwork.com/defense-news/2019/07/upcoming-policy-shaping-dods-newly-renamed-security-clearance-agency/. Accessed May 28, 2021.
Polantz, Katelyn.
“Alleged Oath Keeper Charged in Capitol Riot Says He Once Worked for FBI and Holds Security Clearance.”
Last modified
February
9
,
2021
. https://edition.cnn.com/2021/02/09/politics/oath-keeper-capitol-riot-fbi-security-clearance/index.html. Accessed March 10, 2021.
Roblek, Vasja, Mirjana Pejić Bach, Maja Meško, and Tine Bertonce.
“Best Practices of the Social Innovations in the Framework of the E-Government Evolution.”
Amfiteatru Economic
22
, no.
53
(
2020
):
275
302
. https://www.amfiteatrueconomic.ro/temp/Article_2868.pdf.
Schreier, Margrit. “Qualitative Content Analysis.” In
The SAGE Handbook of Qualitative Data Analysis
, edited by Uwe Flick,
170
84
, 176.
London
:
SAGE Publications
,
2014
.
Shepherd, Katie, John Hudson, and Spencer S. Hsu.
“State Department Aide Appointed by Trump Stormed the Capitol, Beat Police with a Riot Shield, FBI Says.”
Last modified
March
5
,
2021
. https://www.washingtonpost.com/nation/2021/03/05/trump-federico-klein-capitol-riot/. Accessed March 9, 2021.
Reuters
.
“Crisis at the Capitol.”
Last modified
January
7
,
2021
. https://graphics.reuters.com/USA-ELECTION/PROTESTS/qmyvmqewmvr/. Accessed March 10, 2021.
US Congress, House
.
“H.R. 490 - Security Clearance Reform Act of 2015.”
114th Congress, 1st Session,
January
22
,
2015
. https://www.congress.gov/bill/114th-congress/house-bill/490/text.
US Congress, House
.
“The Insider Threat to Homeland Security: Examining Our Nation’s Security Clearance Process.”
113th Congress, 1st Session,
November
13
,
2013
, 43. https://www.hsdl.org/?view&did=755448.
US Congress, House, Committee on Oversight and Government Reform
.
“DC Navy Yard Shooting: Fixing the Security Clearance Process.”
113th Congress, 2nd Session,
February
11
,
2014
. https://www.govinfo.gov/content/pkg/CHRG-113hhrg87892/html/CHRG-113hhrg87892.htm.
US Congress, House, Permanent Select Committee on Intelligence
.
“Committees Open Review of Insurrection at Capitol and Threats to Stop Peaceful Transition of Power.”
Last modified
January
16
,
2021
. https://intelligence.house.gov/news/documentsingle.aspx?DocumentID=1110. Accessed March 10, 2021.
US Congress, House, Subcommittee on Oversight and Management Efficiency of the Committee on Homeland Security
.
“Facility Protection: Implications of the Navy Yard Shooting on National Security.”
113th Congress, 1st Session,
October
30
,
2013
. https://www.govinfo.gov/content/pkg/CHRG-113hhrg87184/pdf/CHRG-113hhrg87184.pdf.
US Congress, Senate
.
“Safeguarding Our Nation’s Secrets: Examining the Security Clearance Process.”
113th. Congress, 1st Session,
November
20
,
2013
. Accessed July 16, 2020. https://www.govinfo.gov/content/pkg/CHRG-113shrg86638/html/CHRG-113shrg86638.htm.
US Congress, Senate
.
“S. 113-283 - Enhanced Security Clearance Act of 2014.”
113th Congress,
December
2
,
2014
. https://www.congress.gov/congressional-report/113th-congress/senate-report/283/1.
US Congress, Senate, Committee on Homeland Security and Governmental Affairs
.
“Security Clearance Reform: Sustaining Progress for the Future.”
112th Congress, 2nd Session,
June
21
,
2012
. https://www.govinfo.gov/content/pkg/CHRG-112shrg75219/pdf/CHRG-112shrg75219.pdf.
US Congress, Senate, Committee on Homeland Security and Governmental Affairs
.
“The Navy Yard Tragedy.”
113th Congress, 1st Session,
October
31
,
2013
, 118. https://www.govinfo.gov/app/details/CHRG-113shrg85500/context.
US Department of Defense – Defense Security Service
.
“Periodic Reinvestigations.”
Internet Archive Wayback Machine. Last accessed
March
15
,
2021
. https://web.archive.org/web/20181209165428/http://www.dss.mil/psmo-i/indus_psmo-i_updates.html. Access March 15, 2021.
US Federal Bureau of Investigation
.
“Most Wanted: U.S. Capitol Violence.”
Last accessed
April
25
,
2021
. https://www.fbi.gov/wanted/capitol-violence#Seeking-Info. Access April 25, 2021.
US Government Accountability Office
.
“GAO’s 2005 High-Risk Update, 2005.”
Report no. GAO-05-350T. Last modified
Feb
17
,
2005
. https://www.gao.gov/products/GAO-05-350T.
US Government Accountability Office
.
“Personnel Security Clearances: Additional Guidance and Oversight Needed at DHS and DOD to Ensure Consistent Application of Revocation Process.”
Report no. GAO 14-640. Last modified
September
2014
. https://www.gao.gov/assets/gao-14-640.pdf.
US Government Accountability Office
.
“Personnel Security Clearances: Plans Needed to Fully Implement and Oversee Continuous Evaluation of Clearance Holders.”
Report no. GAO-18-117. Last modified
November
21
,
2017
, 8. Accessed July 16, 2020. https://www.gao.gov/products/GAO-18-117.
US Government Accountability Office
.
“Security Clearances: Additional Mechanisms May Aid Federal Tax-Debt Detection.”
GAO13-733. Last Modified
October
31
,
2013
. Accessed July 16, 2020. https://www.gao.gov/products/GAO-13-733.
US Government Accountability Office
.
“Substantial Efforts Needed to Achieve Greater Progress on High-Risk Areas.”
Report no. GAO-19-157SP.
March
2019
. Accessed July 16, 2020. https://www.gao.gov/assets/700/697245.pdf.
US Office of Personnel Management
.
“Introduction of Credentialing, Suitability, and Security Clearance Decision-Making Guide.”
January
14
,
2008
, 10. Accessed March 13, 2021. https://www.opm.gov/suitability/suitability-executive-agent/policy/decision-making-guide.pdf.
US Office of Personnel Management
.
“OPM Prepares for Transfer of the Department of Defense’s Background Investigations Services Division.”
Last modified
January
31
,
2005
. https://www.opm.gov/news/releases/2005/01/opm-prepares-for-transfer-of-the-department-of-defenses-background-investigations-services-division/.
W., D.
“Modernizing the Security Clearance Process through Machine Learning and AI.”
Last modified
May
30
,
2019
. https://luminaanalytics.com/2019/05/30/modernizing-the- security-clearance-process-through-machine-learning-and-ai/.
Winter, Susan J. and S. Lynne Taylor. “The Role of Information Technology in the Transformation of Work: A Comparison of Post-Industrial, Industrial, and Proto-Industrial Organization.” In
Information Technology and Organizational Transformation: History, Rhetoric and Preface
, edited by JoAnne Yates and Joahn Van Maanen,
7
34
.
Thousand Oaks, CA
:
Sage
,
2000
.