Bad Watchdog Season 2 launches June 20.

Championing Responsible National Security Policy

Insurrection’s Eve

Inside the Capitol Police’s Intelligence Dysfunction Leading Up to January 6
(Illustration: Renzo Velez / POGO; Photos: Getty Images; Lorie Shaull (CC BY-SA 2.0))

In the weeks before January 6, 2021, then-newly installed officials in charge of the Capitol Police’s intelligence division made changes that hampered its ability to provide earlier, more comprehensive, consistent, and prominent warnings about the threats that day. Those changes created friction, turmoil, and a significant diversion of scarce intelligence resources at the worst possible time, a new Project On Government Oversight (POGO) investigation shows.

POGO and other outlets reported that Capitol Police personnel, including former Chief Steven Sund and four former analysts, have alleged that mismanagement of the intelligence division contributed to the failure to adequately defend the Capitol. It’s a charge that Capitol Police intelligence division leaders have broadly denied. However, transcripts of interviews with Capitol Police officials by the House select committee that examined the attack, as well as recent congressional and watchdog reports and other records, confirm key claims by four former Capitol Police intelligence analysts. They spoke with POGO about the role mismanagement played in the failure to properly prepare for January 6, specifically the role of the intelligence division’s then-director Jack Donohue and assistant director Julie Farnam.

POGO’s new investigation reveals that in the run-up to January 6, Donohue and Farnam made a number of confounding and counterproductive moves, such as

  • directing their staff to end proactive searches — unless explicitly assigned to them — of public websites for concerning information indicating threats to Congress (Farnam told the select committee that they should have been conducting these proactive searches);
  • removing a team lead with significant intelligence experience who had previously coordinated the division’s assessments of demonstrations at the Capitol (after January 6, the Capitol Police’s watchdog recommended the division have such a team lead);
  • putting analysts into jobs for which they were untrained or had little to no experience (Donohue said, “we deliberately tasked people outside of their comfort zones”);
  • rotating the responsibility for writing an assessment of January 6 threats from one analyst to another in late December 2020 while withholding the first analysts’ work from the second in order “to see what she could come up with,” as Farnam put it;
  • sidelining about a quarter of the intelligence division’s analysts in the immediate days before January 6 by directing them to research the global history of political assassinations and vandalism of local, state, and federal officials’ private property; and
  • mandating a two-hour staff-wide training on the afternoon of January 5, 2021, widely seen as not useful by the analysts, instead of using that time to scour social media for threats the next day.

These changes took place in the context of trying to improve staff training, to harness private sector intelligence tools, and to bolster supervisory oversight. There is no evidence of a malicious effort to sidetrack the intelligence division. But the timing and rollout of the changes, along with personnel tensions, created problems.

“Unfortunately, in the middle of making these major changes, we had an insurrection,” Farnam said in an interview with the select committee. During that interview, she largely denied that the changes hindered the intelligence division’s January 6-related performance. Instead, she and Donohue have fiercely defended what they did. “The changes were necessary,” she told the select committee. “It was a failing team. It was an underperforming team … people didn’t want to get on board with the changes, and that was to the detriment of the division.” And she has claimed on national television that the intelligence division did “an excellent job” despite official findings to the contrary.

It’s impossible to conclusively determine how much these changes might have affected law enforcement preparation on the day thousands of then-President Donald Trump’s supporters stormed the Capitol. But scattered throughout the growing official public record is evidence that their management had adverse effects at a critical time.

Part of that official record is a body of interview transcripts from the House Select Committee to Investigate the January 6th Attack on the United States Capitol, made public months ago, which shed light on these breakdowns. But the select committee’s final report did not highlight that material; instead, it focused on the overriding task of detailing Trump’s anti-democratic efforts to stay in office. That’s been a decision some have criticized in part because the debate over law enforcement and intelligence shortcomings remains fierce, perhaps nowhere more so than among those who worked for the Capitol Police that day — the only agency with the singular mission of defending Congress.

Notably, former Chief Sund says the leaders running Capitol Police intelligence did not sound the alarm loudly enough before January 6. The “biggest intelligence failure was within my department,” Sund wrote in his book.

A security consulting firm listing Donohue as an employee did not respond to POGO’s request for an interview. Farnam’s attorney referred POGO to the Capitol Police. A Capitol Police spokesperson did not respond to POGO’s detailed queries, but told POGO last week that many of the analysts’ claims “are nonsense and all of them are old.” Previously, Farnam, Donohue, and a Capitol Police spokesperson have denied that intelligence failings are to blame. Farnam has pointedly criticized Sund, tweeting in response to his book that “there was ample intelligence provided beforehand necessary to prepare appropriately. You did not.”

While the select committee’s final report does not resolve disputes such as these, a mounting number of other government reports have found that the Capitol Police intelligence division did fall short in warning about January 6 threats. The latest is from the nonpartisan Government Accountability Office, which found the intelligence division’s leaders didn’t exercise sufficient supervisory review, resulting in inconsistencies in its threat products, failures to include relevant intelligence, and burying “actionable information” deep in its assessments. Those findings are consistent with an earlier bipartisan Senate review and a Capitol Police inspector general report.

The breakdowns in the Capitol Police’s intelligence division are far from the whole story of what went wrong in the lead-up to and on January 6, even within the Capitol Police writ large — including Sund. But it is a vital chapter. Without a full telling, the right lessons may not be learned, setting the stage for another close call or worse if Congress faces another assault.

In principle, the United States Capitol Police (USCP) has said it embraces learning the lessons from January 6. “Everyone who is charged with the safety of this region must study the mistakes of January 6 to ensure an attack like that never happens again,” the Capitol Police has told the press. “We want the public to know the USCP has made dozens of changes to improve intelligence, planning, communication, training, and equipment, among other reforms.”

Despite this stated desire to learn from mistakes, the attitude has been notably less welcoming when it comes to the former analysts, even though key aspects of their claims are backed up by evidence. “Analysts don’t always see everything in the final picture because they are sometimes compartmentalized,” Donohue said. Still, he and Farnam have both conceded there were problems while they were in charge that undermined the intelligence assessments that were meant to make the Capitol Police aware of January 6 threats. A higher-level official overseeing Donohue and Farnam cryptically told the select committee, when he was asked if the changes Donohue and Farnam enacted “disadvantaged” the Capitol Police, that he agreed it’s “safe to say there was a gap” in the lead-up to January 6.

Although the Capitol Police and Farnam’s attorney have denied there was reprisal, the wounds fester to this day.

The divide isn’t just over substance. The former analysts say Farnam retaliated against them subsequent to making after-action reports or making other disclosures that linked her and Donohue’s management of the intelligence division to the Capitol Police’s lack of preparation on January 6. In POGO’s conversations with the former analysts, they have mostly focused on Farnam. They had far more interaction with her because she, unlike Donohue, supervised them on a daily basis. She also started at the Capitol Police shortly before Donohue and continued there after Donohue left the force only a few months following January 6.

Although the Capitol Police and Farnam’s attorney have denied there was reprisal, the wounds fester to this day. One analyst spoke to POGO in his first time going on the record, while the other three spoke on the condition their names would not be revealed in part due to fear of continuing retaliation or of jeopardizing future career prospects.

An Early, Unheeded Warning

Just days after Farnam started in her new role at the Capitol Police in October 2020, and weeks before Donohue came on board, analysts warned her against abruptly making sweeping changes to the intelligence division on the eve of a perilous time for Congress.

According to Farnam, during her first meeting that November with one of the analysts, Farnam said she was told that “I needed to learn more about the team before I tried to make changes.” The analyst was worried at the time that “something is going to fall between the cracks” if there were major, abrupt shifts in how the under-resourced intelligence division operated during the tense time between the 2020 presidential election and the inauguration. (A second analyst confirmed that this analyst was worried about this at the time.)

That conversation was contentious, both parties agree. Farnam told the select committee that the analyst was “belligerent” — an accusation denied by the former analyst, who says that it was Farnam who first expressed anger after the analyst began to tell Farnam what she did in the intelligence division. A core point of contention was that the analyst had been leading a team of analysts. This raised Farnam’s ire, who responded that the analyst was not a supervisor. The former analyst similarly recalled Farnam saying, “I am the only supervisor around here.”

That episode set the stage for heightened tensions. Farnam said, “I ended up writing her up 11 days into my job. So that’s not how I wanted to get started with the team.”

The analyst’s warning went unheeded, and others since then have said the sweeping changes in the intelligence division were poorly timed. “Initiating significant changes in mission, practices and procedures of the intelligence unit in the midst of a pandemic, social political upheaval while home grown anarchists were metastasizing on social media was a recipe for disaster,” Terry Gainer, a former chief of the Capitol Police, told CNN.

Eric Hoar, one of the former analysts, told POGO that even without Trump’s inflammatory lies about the stolen election, the upcoming Joint Session of Congress on January 6 would still have been a reason to hold off on making big changes that could undermine the intelligence division’s effectiveness in the short run. He said that any time both the House and Senate hold a joint session together, it’s like the Super Bowl for the Capitol Police: The stakes are particularly high.

Making matters worse, Farnam sent the team out to play without a quarterback.

That same analyst who Farnam wrote up was also removed as team lead, even though she had previously led assessments of major demonstrations and other special events.

That same analyst who Farnam wrote up was also removed as team lead, even though she had previously led assessments of major demonstrations and other special events. “That experience lets you sift through the noise and pick up what’s important,” one former analyst told POGO. Her removal meant the intelligence division was heading toward the Joint Session to be held on January 6, 2021, without a seasoned person overseeing, synthesizing, and assessing the intelligence coming in. The former analysts said this team lead also would have normally been involved in talking to protest organizers — a job Farnam took over despite having no apparent prior experience in assessing threats from demonstrations.

In the wake of January 6, the Capitol Police’s inspector general recommended that the intelligence division have a “group leader or middle management positions that would expand supervisory coverage” — the kind of role that Farnam had eliminated shortly before the attack on the Capitol.

The analyst Farnam removed as team lead also had well-established ties with Sund and another top official, Sean Gallagher, who was viewed as the main avenue to get intelligence to the entire Capitol Police leadership team, according to the former analysts. “There is no question they [Capitol Police leaders] would have been beaten over the head” with intelligence about threats to the Capitol earlier had the former team lead stayed in her role, a former analyst told POGO. Farnam herself would tell the select committee that her lack of established relationships during this period hampered her ability to break through to Capitol Police leadership. “Me being new, not knowing who the players were,” she said to the select committee, “I didn’t want to ruffle too many feathers.”

“If this happened today,” she said nearly a year after January 6, “I would have been much more vocal and much more adamant about ensuring the correct steps were going to be taken.”

In the weeks after November 2020 building up to that day of infamy, the rift between Farnam and roughly half of the intelligence division would only get wider.

Halting Proactive Open-Source Intel Research

There’s a good case to be made that the storming of the Capitol on January 6, 2021, is the biggest intelligence failure in U.S. history. Unlike in military operations or terrorist plots, thousands of online posts by pro-Trump supporters were public, available to anyone with internet access who was looking. While the heated First Amendment-protected rhetoric on its own may have been too vague to make arrests in advance of the attack, it was widespread and concerning enough that it should have spurred significantly heightened security precautions at the Capitol.

However, early on, Farnam and Donohue curtailed the Capitol Police’s ability to research concerning online posts about January 6. In the national security and law enforcement communities, examining online posts is a major part of open-source intelligence gathering. Open-source intelligence collection includes combing through social media sites and other publicly available sources online for evidence of threats, of which there were many in the weeks before January 6.

Donohue seemed to understand the importance of open-source intelligence gathering. Four months before he started at the Capitol Police, he gave prescient testimony to Congress, warning that “seditionists actively promote a revolution.” He called for supercharging open-source monitoring that would “generate finished intelligence rapidly and at a massive scale that can perceive imminent threats before they emerge.” And he did make some effort to improve the Capitol Police’s open-source collection. On December 16, 2020, he sought to have a social media monitoring company create an alert for “Joint Session of Congress.”

Still, the intelligence division took some backwards steps on open-source intelligence: According to the former analysts, the Capitol Police’s own in-house capabilities were significantly shuttered.

“The biggest and most detrimental change was she [Farnam] disbanded the Open Source Section,” said Hoar, a seasoned intelligence analyst who spent 19 years with the Capitol Police until leaving last year, in an interview with POGO. “Their focus was specifically to collect and analyze open-source intelligence in support of assessments for events, threat assessments, and to support criminal investigations.”

In her interview with the select committee, Farnam confirmed that she had indeed “eliminated” the Open Source Section by combining it with another group under her supervision. “It was such a small team anyways, it made sense,” she said.

The biggest and most detrimental change was she disbanded the Open Source Section.

Eric Hoar, former Capitol Police intelligence analyst

This was not just harmless bureaucratic reshuffling. What Farnam didn’t tell the select committee is that she and Donohue also changed what the analysts in the former Open Source Section focused on in those weeks before January 6: They did not have their analysts focus their time on pursuing proactive open-source searches related to the upcoming Joint Session.

This was at odds with past practice. The Open Source Section’s priorities normally would shift depending on the needs of the Capitol Police, the former analysts told POGO. For instance, supporting criminal investigations would have been less of a priority when a high-stakes event was imminent, such as the Joint Session on January 6. Instead, in the days before January 6, Donohue and Farnam had the analysts prioritize researching a huge backlog of mostly non-January 6-related threats made against individual lawmakers.

Hoar and another former Open Source Section analyst told POGO that the cases they researched before January 6 had nothing to do with that day. “Definitely none of the ones I worked were related to January 6,” said one former analyst. “They were old.”

“Our proactive open-source operations pretty much stopped right there,” Hoar said, “and all we did was answer requests for information from other departments and everything came through Farnam.” The other three former analysts also told POGO, and an email written by Farnam obtained by POGO confirms, that she directed them to only work on that backlog unless she specifically tasked them with assignments.

Donohue, too, directed analysts to prioritize addressing the backlog, even as the events of January 6 were imminent.

At the end of the day on January 4, Donohue emailed analysts to “please bring your ‘A Game’ to our collective intelligence gathering and dissemination responsibilities,” given “we are entering a two-day period of protest activity that has higher than usual potential for serious disruption.”

But that message was at odds with what he directed them to do. “At the same time, we still have an abundance of … cases that must be investigated, please use time before the protests and during less intense periods during” the demonstrations to work through them, he wrote.

Donohue noted there were two analysts exempt from this: One analyst would monitor open source intelligence during the day, and one at night. But given the avalanche of social media activity associated with January 6, that skeleton crew was not nearly enough.

Donohue and Farnam’s approach contrasted with a more productive environment of trust the analysts had under a previous manager. Under that former official, the Open Source Section’s analysts were broadly allowed to take initiative to identify concerning intelligence and threats to the Capitol, and their efforts were coordinated by the team lead, they said.

For instance, they pointed to the proactive approach taken during the racial justice rallies in the summer of 2020. Hoar told POGO that, as a precautionary measure, the Open Source Section’s analysts researched social media for information on Black Lives Matter protests to learn where they were being planned and what they intended. They correctly predicted that there were no serious threats for the Capitol Police to deal with. (Reporters have written about the striking contrast between law enforcement’s excessive use of force against Black Lives Matter protestors at Lafayette Square adjacent to the White House on June 1, 2020, and the lack of preparedness at the Capitol on January 6, 2021 — but the Capitol Police were not part of the Lafayette Square episode.)

When Farnam was interviewed by the select committee nearly a year after the January 6 insurrection, she did acknowledge that the paucity of proactive open-source searching was a problem. However, she failed to disclose that she and Donohue directed analysts to spend their time on addressing the backlog unless explicitly told to do otherwise.

This directive set Farnam up to be a single point of failure. And in the wake of her tense first conversation with the former Open Source Section team lead in November that led to a written reprimand, the analysts were wary of crossing her.

“They should be proactively seeking out information, information about threats, and identifying those threats proactively,” Farnam said of the intelligence division. “We weren’t doing that.”

“There were a couple searches that we did, but they were assigned out to them,” she said.

As she was criticizing her division for not proactively combing through social media and online posts, a select committee staffer asked who made these assignments. Farnam responded, “I did the assigning.”

In his book, former Chief Sund wrote, “the Open Source Section, one of the most critical assets for monitoring demonstrations and online threats against the Capitol and members of Congress, was decimated, with many analysts being directed to clear backlogs of assignments for investigations instead of monitoring open-source information.”

“We Deliberately Tasked People Outside of Their Comfort Zones”

According to Donohue and Farnam, one of their goals early on was to take stock of what each of the analysts could do in order to identify gaps in their training and create a more robust training program. But this worthy effort was poorly timed and implemented, and it adversely affected the quality of the intelligence division’s assessments of January 6 threats.

During his interview with the select committee, Donohue said they shuffled people around on purpose. He said that, as part of an effort to assess the analysts, “we deliberately tasked people outside of their comfort zones to produce different types of reports or to engage in different type [sic] of analyses or to leverage some of the technologies that they may not have been comfortable with.”

But it was a terrible time to poorly utilize the scarce staff resources within the Capitol Police’s intelligence division.

For example, instead of having one person consistently leading the monitoring and analysis of intelligence in the run-up to January 6, as would normally occur, Farnam rotated that responsibility to different analysts, who each then had to quickly get up to speed.

When rotating that job on Monday, December 21, 2020, Farnam did not share the work of the first analyst with a second one because Farnam “wanted to see what she [the second analyst] came up with on her own.”

The second analyst, who worked a Tuesday through Saturday schedule, saw the email when she got into work. It had no deadline. But on Wednesday, December 23, Farnam pressed to have the assessment by lunch, which left the second analyst scrambling to finish. She had only about 12 hours total to conduct research from scratch and write up a January 6 threat assessment, which she told House Republicans would normally take much longer.

Under the conditions created by Farnam and Donohue, in which no analyst was consistently leading the intelligence division’s assessment of January 6 threats, no one in the division had enough time to both stay on top of the intelligence and to fully synthesize the meaning of new information coming in, the former analysts told POGO. This period was key, as just days earlier on December 19, Trump tweeted to his supporters to come to DC on January 6: “Be there, will be wild!” That tweet supercharged the potential threat and led to an explosion in social media posts. It was only briefly mentioned in the December 23 assessment.

After receiving the assessment, which was produced under rushed conditions, Farnam said during her interview with the select committee that the analyst “had a lot of good information, but there was no analysis.” Although Farnam criticized the analyst’s work, she did not share with the select committee the conditions under which that assessment was produced.

“The Capitol Was Surrounded”

There were other consequences of not having one intelligence analyst consistently in charge of assessing the demonstrations being planned and the potential for violence in advance of January 6: No one had a bird’s-eye view. That bigger view could have allowed the Capitol Police to better understand how many of the groups seeking protest permits were potentially fronts for Stop the Steal organizers and could have allowed the police to take proactive steps that would have improved the defense of the Capitol complex. This matters because a much larger, single protest can pose a more serious threat than numerous, small demonstrations. Moreover, suspected efforts by January 6 organizers to conceal the true nature of the size of their demonstration by using proxy groups to seek permits is an additional red flag that should have led to better preparation.

The Capitol was surrounded by six different permitted groups that day that gave them incredible tactical advantage.

A former Capitol Police intelligence analyst

“The Capitol was surrounded by six different permitted groups that day that gave them incredible tactical advantage,” said a former analyst. But that information was siloed among the analysts. “We didn’t know that there were other permit assessments [other than those each analyst handled] because of the way Julie [Farnam] was randomly assigning work and there was no central log.”

“Had we had that information and spelled that out, we could have denied those permits,” the former analyst said.

While it’s unprecedented for the Capitol Police to deny permits, the former analyst said the Capitol Police could have also moved the permitted areas so that police could have more easily managed crowds.

The Capitol Police’s special events staff that received requests for permits “told Julie these groups are filing by different names, but it’s the same woman hanging up and calling back again,” the former analyst said.

The former analyst could not remember the woman’s name but said it could be Kylie Kremer, whom text messages from January 6 event organizers show was trying to obtain permits for multiple groups. In a text exchange on January 4, Kremer told prominent Trump supporter and election fraud conspiracy theorist Mike Lindell she had foreknowledge that Trump would “unexpectedly” call for a march from the Ellipse to the Capitol or the Supreme Court on January 6. “This stays only between us,” she wrote.

The special events staff flagged concerns that all of the groups were potential proxies for Stop the Steal, records obtained by BuzzFeed show.

It was a potential moment for intervention — and more thorough intelligence analysis could have provided more evidence that the Capitol Police could face trouble defending Congress if the groups were Stop the Steal proxies.

The former analyst said that if the assessments had been handled like they were before Farnam and Donohue’s arrival, the concerns about all the proxy groups would have gone to senior leadership “so they could see that that was going to be a problem.”

It turns out that Farnam did raise some concern about this, but for only a couple of the groups. On December 31, 2020, she emailed Capitol Police leadership that two groups could be proxies for Stop the Steal. But it seems she conducted no follow-up. “I don’t know what the discussions were once it got sent forward,” Farnam told the select committee nearly a year later.

Although she emailed top officials, Farnam also believed the groups requesting the permits would adhere to the 50 person per demonstration limits in place due to the pandemic. A House Republican report says that Farnam testified to them as much. That report criticized her because, over the course of 2020, there had been “massive protests throughout the country where ‘COVID protocols’ were ignored.”

It was strange to assume only 300 protestors total would be present at any one time across the six permitted Capitol sites. Rally organizers had themselves told multiple agencies in December 2020 that potentially tens of thousands of protestors were coming to DC on January 6. DC’s Metropolitan Police Department denied protest permits on that basis, but did not have authority over the land under the Capitol Police’s jurisdiction. Yet Farnam and Donohue both told the select committee they were in contact with the Metropolitan Police, including during conversations with protest organizers. It’s unclear whether Farnam and Donohue pressed the organizers on the numbers of people that might show up. Prior to January 6, law enforcement estimated up to 80,000 people could be demonstrating, according to the then-Army secretary.

Records show that, despite special event staff concerns about the potential proxies, the permits were ultimately approved. Their permits were greenlit in part because of a “lack of evidence” that each demonstration would have more than 50 people. A top Capitol Police official also told the select committee that the concerns Farnam raised about some of the groups prompted discussion but no action. “The result was, based on what was provided [Farnam’s concerns], no, that wouldn’t be sufficient to deny the permit,” the official said.

The Diversion of Significant Staff Resources

Given Trump’s tweet to his supporters calling on them to protest in DC on the day Congress was set to certify Joe Biden’s win and the widespread online posts calling for violence, it was clear then that the limited resources of the Capitol Police’s intelligence division needed to be well-utilized. They were not.

Donohue told the select committee that the intelligence division needed a lot more analysts to tackle its immense annual workload. He also said that, two days before January 6, “I had never seen the amount or frequency in volume of this sort of chatter since the rise of the ISIS caliphate.” Confoundingly, in the midst of the staggering number of concerning statements being made about January 6, he and Farnam tasked many of their analysts — the intelligence division had only about a dozen analysts at the time — with massive research projects that seemed to be of limited practical use in that moment.

For instance, just days before January 6, several analysts were tasked with producing research on the history of political assassinations from around the world, and on vandalism of local, state, and federal officials’ private property.

“There’s a lot of countries in the world. Assassinations have happened for centuries, if not thousands of years. But so that’s what we were working on, three analysts leading up to January 6,” a former analyst said.

When asked about this by House Republicans, Farnam confirmed she assigned these research projects. “One of the projects was to look at assassinations and assassination attempts against elected officials because we have seen a huge surge in threats,” she said, according to a House Republican report.

More broadly, instead of supervising a team of intelligence analysts, Farnam and Donohue took on conducting the analysis themselves and failed to harness the division’s collective resources. “I felt coming on board I had to do a lot of the analytic work myself. I’m not going to speak for Jack [Donahue],” Farnam told House Republicans, “but I know the two of us did a lot of, like, the analyst type work in the beginning because the team didn’t have the capability.”

Farnam and Donohue, the former analysts said, had an erroneous belief that many of their team members lacked adequate training. One former analyst told POGO she has gone through the FBI academy and been through CIA and National Security Agency trainings. “They’re trying to portray us as a bunch of untrained people and that’s so far from the truth,” the former analyst said. Hoar similarly has extensive training. They both said that they were among those analysts whose skills were not sufficiently used during this key time.

I had never seen the amount or frequency in volume of this sort of chatter since the rise of the ISIS caliphate.

John Donohue, former Capitol Police intelligence director

Fueling Farnam’s and Donohue’s impressions was that they had put people into roles they were largely untrained to perform and had given people unrealistically short amounts of time to perform analysis, the former analysts said. Those actions contributed to the problems Farnam and Donohue would later cite as intelligence office shortcomings.

Still, there were a few analysts in the intelligence division who did lack adequate training, and overall there wasn’t a robust, structured program in place to rectify that problem. The Capitol Police inspector general found that there were “intelligence-related deficiencies” that included inadequate training and professional standards. “A formal Intelligence Training Program is a must,” states an inspector general report from March 2021.

Although launching a training program was an important priority, it paled in comparison to the more pressing need of marshalling the limited resources of the intelligence division to identify threats and prepare the Capitol Police immediately before January 6. Given that, one of the bigger head-scratchers in terms of timing is, on January 5, 2021, Farnam arranged for the analysts to attend a two-hour training that afternoon which provided “an overview of how to conduct open source research.” The former analysts all said the training was not especially useful.

“Why were we not all focused on scouring social media to find as much information as we could the day before?” said a former analyst.

Mixed Messages from Capitol Police’s Intelligence Division

Without question, the Capitol Police were unprepared on January 6 for the onslaught they would face. But the final report from the House select committee on the January 6 attack leans against blaming intelligence. It found that “the intelligence community and law enforcement agencies detected the planning for potential violence directed at the joint session of Congress.” At an event at the University of Virginia in March 2023, the committee’s chief investigative counsel Tim Heaphy said it was “a failure to operationalize that intelligence” — nearly identical language Farnam used in her select committee interview. During that interview, she said, “I don’t think it was a failure of intelligence; I think it was a failure to operationalize the intelligence.”

While the report avoids laying explicit blame on any individuals for law enforcement shortcomings, it highlights questionable decisions by then-Capitol Police Chief Steven Sund, such as ultimately rejecting requesting the National Guard in advance, not ensuring there was a department-wide plan in place, and not canceling his officers’ days off despite saying the Capitol Police had an “all hands on deck” posture.

The Capitol Police did take some actions. Then-Deputy Chief Sean Gallagher, who reported to Sund, told select committee staff that there were some improvements like beefing up protection of lawmakers, posting officers at Capitol tunnel entrances, placing evacuation vehicles in key positions, and pushing bike racks out to create a larger perimeter around the Capitol. But these changes were not enough.

Despite the apparent failure by Sund to act sufficiently on intelligence, there’s more to the story — and it’s mostly missing from the select committee’s report.

Despite the apparent failure by Sund to act sufficiently on intelligence, there’s more to the story — and it’s mostly missing from the select committee’s report.

The intelligence Sund did not act sufficiently on is a January 3, 2021, special assessment that contained a strong warning in one paragraph. “Supporters of the current president see January 6, 2021, as the last opportunity to overturn the results of the presidential election. This sense of desperation and disappointment may lead to more of an incentive to become violent,” according to part of the paragraph. “Unlike previous post-election protests, the targets of the pro-Trump supporters are not necessarily the counter-protestors as they were previously, but rather Congress itself is the target on the 6th.” The paragraph provides further troubling information, such as “Stop the Steal’s propensity to attract white supremacists, militia members, and others who actively promote violence, may lead to a significantly dangerous situation for law enforcement and the general public alike.”

Unfortunately, the paragraph with this warning is on page 13 of the 15-page assessment and is contradicted by passages earlier in the assessment.

A bipartisan report produced by two Senate committees in the summer of 2021 and the Capitol Police inspector general found that the January 3 assessment’s strong warning near the end on page 13 was at odds with the assessment’s “bottom line up front,” or BLUF, on its first page.

“Inconsistencies between intelligence products, and within the January 3 Special Assessment, led to a lack of consensus about the gravity of the threat posed on January 6, 2021,” the Senate review found. “The BLUF does not adequately summarize the analyst’s findings,” stated the Capitol Police inspector general, and, as a result, a reader “could draw an inaccurate conclusion” based on it.

During her interview with the select committee, Farnam was asked if the language on page 13 should have been in the BLUF itself.

“I think now it would be. Or we would take pieces of that. Like, Congress is going to be targeted, that might be important to put in the BLUF,” she said.

However, she then defended herself with the inaccurate statement, “But, at the time, we weren’t doing BLUFs at all.”

In striking contrast to the paragraph on page 13 of the January 3 assessment, on page 2 it stated that “the protests/rallies are expected to be similar to the previous Million MAGA March rallies in November and December 2020, which drew tens of thousands of participants.” This comparison might have given the impression that there was not much of a threat to the Capitol, given that those earlier rallies primarily involved areas outside the Capitol grounds and did not result in any large-scale or sustained assault on a government building with the goal of getting to the officials within it; rather, the main law enforcement challenge was rally participants’ skirmishes with counter-protesters.

Select committee staff pointed Farnam to that language and asked her if that might “undercut … your later warning that this third demonstration might be different from the previous two.” She responded, “in retrospect, I would’ve rephrased that.”

Then the select committee staff showed her the first page of the Capitol Police’s Civil Disturbance Unit’s operational plan for January 6, which stated that day would be like the earlier protests. She agreed that the Capitol Police may have based their planning on the language on page 2 of the assessment.

There are other confusing aspects of the assessment as well — and nearly all of it sends a far less disturbing signal earlier in the assessment than the warning at the very end.

On page 3, it states that “no groups are expected to march and all are planning to stay in their designated areas,” and on page 5 it states that “the number of people who indicate they are going to the event listed on these social media postings is relatively low.” But on page 8 it states that “participants will gather at the Ellipse and march to the U.S. Capitol in support of President Trump.”

The Senate review further found that the Capitol Police produced three daily intelligence reports after the January 3 assessment that also played down any risk of violence on January 6. Those mixed messages are due in part to the failure of Donohue and Farnam to adequately supervise the work of the analysts in the intelligence division. The Senate report says those subsequent reports went out “without supervisory review” — a finding recently echoed by the nonpartisan Government Accountability Office. Farnam would tell House Republicans she should “have given those reports more attention on January 4, 5, and 6.”

The former analysts say the problems within the January 3 assessment and the broader inconsistencies could have been avoided had Donohue and Farnam properly utilized the division’s resources during the preceding weeks, focused their time on reviewing the products going out the door, and ensured that leadership and Capitol Police as a whole had the information they needed. Their haphazard approach is further illustrated by comments Farnam made to the select committee. She told them she wrote the warning on page 13 of the January 3 assessment in “literally less than 5 minutes” while in bed. “There wasn’t a whole lot of, like, analysis and thought, like let me wordsmith this,” she told the select committee. Unfortunately, she and Donohue did not devote the time to ensuring the full assessment and subsequent intelligence products were consistent with that one-paragraph warning.

Intelligence missteps cascaded into inadequate preparation, which placed the health and lives of front-line officers at risk.

Tim Blodgett, the House’s deputy sergeant at arms

As the Senate review found, the Capitol Police intelligence division “possessed information about the potential for violence at the Capitol on January 6 but did not convey the full scope of information, which affected its preparations.”

“Intelligence missteps cascaded into inadequate preparation, which placed the health and lives of front-line officers at risk,” Tim Blodgett, the House’s then-acting sergeant at arms, testified to Congress in February 2021. “Warnings should not be qualified or hidden. Bad information, conflicting information or missing information leads to poor decisions.”

In the Aftermath, Reprisal

In the wake of the attack on the Capitol, several analysts began to raise questions about the intelligence division’s performance, file critical after-action reports, and make disclosures to Congress and to their inspector general. The four former analysts POGO interviewed said they were met with reprisal principally from Farnam, who, along with Donohue, was the main focus of their whistleblowing.

“We were retaliated against unmercifully,” one former analyst told POGO. Many of them said the retaliation they faced came on top of the PTSD and depression that was fueled by having to watch over and over again the videos of their colleagues being relentlessly assaulted by the crowd, as they helped law enforcement identify and build cases against persons accused of crimes committed that day.

The timing of at least one of the actions taken by Farnam against Hoar is particularly suspect, and there is evidence she had animus against him and the others.

We were retaliated against unmercifully.

A former Capitol Police intelligence analyst

On the morning of Friday, December 10, 2021, Hoar participated in an interview with the select committee that was organized through the Capitol Police’s Office of General Counsel. He had prepared in advance by ensuring he would be able to finish his work assignments due that day. “When I returned from the interview, Farnam piled on several extra assignments with only a couple hours until the end of the day,” Hoar told POGO. On the following Monday, she issued him an administrative reprimand for mismanagement of his time because he did not finish all of these additional assignments.

Her own interview with the select committee would come a few days later, during which she identified him by name as one of the analysts “who outright refused to get on board with the changes.” She also denied criticisms he made in a January 9, 2021, email that CNN had reported on just weeks before her interview.

This was far from the only time Farnam issued Hoar an administrative reprimand, Hoar told POGO. He said that after he filed a complaint about her with the Capitol Police’s inspector general she began issuing him weekly reprimands “for almost a year, until the work environment was so hostile I had no choice but to resign and take a job elsewhere.” He said that in the 18 years prior to his January 6-related disclosures, including months under her supervision, he never once received an administrative write-up for poor work performance or any other issue.

Farnam and the Capitol Police have denied retaliating against the former analysts. They cite the Capitol Police inspector general’s findings that there was no reprisal. However, a lawmaker alleged in January 2022 that the inspector general’s reprisal review was a “check-the-box investigation.” It is unclear what allegations and evidence the inspector general considered, as those investigative reports are not public, and no details were provided to the analysts aside from a brief, one-page letter from February 2022 informing them that the matter was closed.

Capitol Police whistleblowers face a deck that’s particularly stacked against them. Attorneys say that the law protecting Capitol Police whistleblowers from retaliation is weaker than protections for most other federal employees.

A “Colossal Intelligence Failure,” Bad Decision-Making, or Both?

After January 6, the Capitol Police’s leadership came under intense press and congressional scrutiny. Chief Sund left his position on January 8, after then-Speaker of the House Nancy Pelosi called for his resignation.

Sund has said others were at fault, and has blamed a “colossal intelligence failure” for January 6.

When asked by select committee staff to explain what he meant by that, Sund said “the synthesis and presentment of intelligence in IICD [Intelligence and Interagency Coordination Division] products and briefings for our use in operational planning and decision making … that’s where I think the biggest issue was.”

Reacting to Sund’s explanation, a staffer said, “It sounds like it’s not a failure of getting the intel; it’s more of a failure of what was done with the intel generally?”

Breakdowns can happen at numerous points from collection of information, to producing and communicating intelligence analysis, to the decision-making based on the analysis. An intelligence analysis could assign too much or too little significance to individual pieces of information, or a key missing piece of the puzzle can skew the overall picture. Conclusions can be poorly communicated. Inadequate sharing of intelligence can render even the best analysis moot. A decision-maker could ignore intelligence or interpret it in a way that wasn’t intended by its authors, even if their message is clear. Biases of various types can creep in throughout, including a failure to imagine certain scenarios even if the available evidence points toward them.

The select committee’s conclusions speak to the front-end and back-end of this process: Several law enforcement agencies collected a lot of raw intelligence pointing to the potential for violence and even an assault on Congress, yet that intelligence wasn’t translated into sufficient law enforcement planning, preparation, or resources on January 6.

However, if the intelligence isn’t communicated well, it can be misunderstood or may not fully reach decision-makers. Sund points to such shortcomings. In his book, he wrote, “There were significant threats to storm buildings and harm members of Congress and other prominent government officials, yet my intelligence leadership wasn’t making this knowledge available to us to assist in our planning or to our stakeholders to support critical decision-making. The IICD had this information for weeks.”

“It could have been a game changer for our preparedness on January 6. Instead, we received a single inconclusive paragraph riddled with qualifiers at the end of a fifteen-page document, a little over forty-eight hours before the day of the event,” wrote Sund.

Despite Sund’s importance, the select committee’s final report doesn’t assess his critique.

Still, many of the select committee’s interviews do offer insights, revealing — when viewed along with the former analysts’ accounts and other information — a disturbing picture of the Capitol Police intelligence shop’s failures in the run-up to January 6.

One failure the transcripts of the committee’s interviews reveal is the lack of imagination on the part of Capitol Police leadership, both in and outside of the intelligence division. Sund, for instance, said he failed to contemplate the scenario that unfolded on January 6. “Having your building surrounded by thousands of people trying to gain access,” Sund told select committee staff, “up until January 6th, that’s not something I really conceived.” When asked if he had been aware of state capitols that had been breached by armed groups, Sund said he was, “but I don’t know how much it played a role in driving my planning.”

Still, many of the select committee’s interviews do offer insights, revealing — when viewed along with the former analysts’ accounts and other information — a disturbing picture of the Capitol Police intelligence shop’s failures in the run-up to January 6.

Press critic Dan Froomkin has written that Sund’s decision-making may have been influenced by implicit bias: He didn’t really think the mostly white crowd of Trump supporters could pose a severe threat. President-elect Joe Biden, the then-head of the DC National Guard, and others have said that they believed a majority Black crowd would have been treated very differently on January 6.

The select committee’s final report does not address Sund’s or other law enforcement’s possible bias. Fifteen of the select committee’s staff members spoke to the Washington Post about their frustration that then-Representative Liz Cheney (R-WY), the select committee’s vice chair, kept the report’s scope narrowly focused on Trump to the exclusion of a number of important aspects of their investigation. Much of what was excluded related to law enforcement and intelligence failures. Questions asked in interviews about the possible role of disparate treatment of Black Lives Matter protestors compared to pro-Trump supporters, as well as remarks by Cheney’s spokesperson, suggest that some of the content the staffers wanted in the final report may have explored the possible role of implicit bias.

“Some staff have submitted subpar material for the report that reflects long-held liberal biases about federal law enforcement, Republicans, and sociological issues outside the scope of the Select Committee’s work,” Cheney’s spokesperson told the Post. “She won’t sign onto any ‘narrative’ that … smears men and women in law enforcement.”

However, there’s been at least one internal account that indicates implicit bias has affected leadership’s assessment of possible threats when protestors were not mostly white. One former unnamed Capitol Police official told ProPublica that even though the intelligence division said Black Lives Matter protests in the summer of 2020 would pose only minimal risks, the response from Capitol Police leadership was, “We don’t trust the intel.”

Although they do not directly implicate Capitol Police leadership’s decisions regarding January 6, there are other reasons bias should be considered. There’s a long history of claims of racism from inside the force. In 1990, 40% of the Capitol Police’s Black officers organized and demanded fair treatment by its leadership; since 2001 over 250 Black Capitol Police officers have sued the agency for racial discrimination (the vast majority are co-plaintiffs in a class action lawsuit that began that year); and there have been other lawsuits filed as recently as fall 2021. Then there’s evidence from January 6 itself: Video from that day shows signs that some Capitol Police officers expressed sympathy with the pro-Trump crowd such as taking a selfie with one of them. The Capitol Police have investigated at least 35 officers for questionable actions taken that day, and have disciplined at least six. A jury convicted one officer for helping a January 6 rioter obstruct justice by recommending he delete Facebook posts to avoid getting caught.

When asked by the select committee “whether the race or the ideology of” pro-Trump protestors meant that they were seen as less of a threat, Sund responded, “No. Always have patrolled and handled myself in an unbiased fashion.”

“You treat everybody the same. But it should be based on intelligence,” Sund said. “Some of the intelligence that they had that should have been incorporated into the products that we were getting, that would have informed a much different response for [sic] the agency, regardless of who they are.”

Whether bias played a role or not, Sund clearly did not fully grasp how the nature of January 6 contrasted with most of the other protests at the Capitol. “Chief Sund said, ‘you know we got big crowds up here all the time,’ failing to appreciate the fundamental difference here,” said the select committee’s chief investigative counsel Heaphy at the March 2023 event at the University of Virginia. Heaphy said the protestors on January 6 saw the members of Congress moving forward to certify the election results as “traitorous.” Capitol Police leadership “just failed to appreciate” that. 

Sund was the only senior Capitol Police leader to take the fall after January 6, but there is evidence others in senior leadership excluded him from at least one important meeting in advance of that day that might have helped him appreciate the severity of the threat. He told the select committee that he “was not invited” to a key intelligence briefing on January 4, 2021, and didn’t learn about the briefing until weeks later. In interviews, Donohue and Farnam separately recounted that when they briefed Capitol Police leadership — with the notable exception of Sund — on January 4, 2021, they were met with “silence” after providing them with a troubling warning.

Sund wasn’t the only one in the Capitol Police who failed to anticipate how bad things would get.

Farnam told the select committee, “I don’t think anyone conceived that they would break in to the Capitol. ... Did I think they would storm the building? Not necessarily.”

There were reasons they should have warned this was a possibility. The intelligence division had numerous online posts in which people were calling on supporters of Trump to “storm” the Capitol, access its tunnels and confront lawmakers in them, and to kill them.

We didn’t say they were going to storm the Capitol, and in retrospect, we probably should have included that.

Julie Farnam, the Capitol Police’s assistant director of intelligence

“We didn’t say they were going to storm the Capitol, and in retrospect, we probably should have included that, but we did say it was going to be violent,” Farnam told the select committee.

When Donohue was asked about this, he acknowledged that people were promoting the storming of the Capitol online prior to January 6, but he avoided directly answering whether he took that possibility seriously even though, for instance, records show that a DC Metropolitan Police Department officer had sent him a tip on January 1, 2021, with a “reference to storming the Capitol.” He also could not recall if he shared any of the concerning raw intelligence with Sund. Instead, he said Trump’s speech at the Ellipse on January 6 “really is the one piece of the puzzle that was what caused it, in my mind.”

Donohue’s argument dovetails with the Capitol Police’s spontaneity defense. “The Department expected and planned for violence from some protesters with ties to domestic terrorist organizations,” the Capitol Police has said in a statement, “but nobody in the law enforcement or intelligence communities imagined, on top of that threat, Americans who were not affiliated with those groups would cause the mayhem to metastasize to a volume uncontrollable for any single law enforcement agency.”

It’s tough to fully square Donohue’s and the Capitol Police’s arguments with the warning buried on page 13 of the January 3 assessment, which seems to have anticipated the possibility that the larger body of protesters, not just isolated extremists, could turn violent.

Of course, that one-paragraph warning only came after numerous pages indicating that January 6 would be like earlier protests.

For its part, the final report of the select committee said in its executive summary that “Capitol Police leadership did not anticipate the scale of the violence that would ensue after President Trump instructed tens of thousands of his supporters in the Ellipse crowd to march to the Capitol, and then tweeted at 2:24 p.m.”

The final report continued, “No intelligence community advance analysis predicted exactly how President Trump would behave; no such analysis recognized the full scale and extent of the threat to the Capitol on January 6th. … Such agencies apparently did not (and potentially could not) anticipate the provocation President Trump would offer the crowd in his Ellipse speech.”

Was the scale of violence impossible to anticipate solely because of the Trump factor? Or was it a failure of leaders like Sund to be open to the possibility, despite available evidence, as Heaphy and Froomkin contend? Or did others, including those in charge of the Capitol Police intelligence, fail to adequately assess or communicate available intelligence to people like Sund, as he claims?

The answers to date are muddled. And all three are plausible, supported by official government reports, and not mutually exclusive.

It’s unreasonable to expect perfect intelligence. But could things have turned out differently even if Trump’s specific actions on January 6 were not anticipated? Had the Capitol Police intelligence division been more effective and issued clearer warnings, would Sund and other leaders have ensured the Capitol Police were better prepared? That’s what Sund and the former analysts contend. It’s a reasonable inference, if impossible to definitively prove.

And assuming there would have been better preparation, would that have been enough to prevent the storming of the Capitol? It depends. If the Capitol Police had real fencing instead of bike racks to help control the crowd, and if substantial reinforcements like the National Guard had been on hand, the day may have turned out quite differently.

What is clear is that Donohue and Farnam mismanaged scarce intelligence resources at a critical time, and the intelligence division under their watch issued a number of mixed and confusing messages about the risk of violence on January 6. They did not clearly, effectively, and consistently communicate the volume and severity of the threats facing the Capitol. There’s a strong case to be made that their management is linked to those failings at the time. Their actions put Congress and their law enforcement colleagues at risk.