Meta description: Fact-checked: what the House Judiciary Committee’s July 2026 markup actually confirmed (44 members’ texts, read directly, not just metadata), what it got right, and which statistics from both parties don’t survive independent verification.
Congress just found out federal investigators secretly read their own text messages. On July 15, 2026, the House Judiciary Committee spent just over four hours marking up seven items, and that discovery (not metadata, actual content) sat behind the most consequential bill of the day. Five of the seven items passed by lopsided, often unanimous votes, with Republicans and Democrats praising each other’s work. The other two passed on near party-line votes of 14-9 and 15-8, and consumed roughly 70% of the session’s runtime. Two of seven bills ate up most of the clock. If you only read a headline that said “committee passes seven bills,” you’d miss both halves of what actually happened: a real, confirmed story about Congress’s own communications being searched, and a bitter fight where neither side’s numbers survive a fact-check fully intact.
Five bills, two fights, one confirmed bombshell: that’s the shape of this House Judiciary Committee markup, and it’s the order this piece follows. The bipartisan wins first, then the confirmed Arctic Frost story, then the two contested items, with every disputed number checked against the record rather than repeated as said.
The Bipartisan Half: Five Bills, Real Cooperation
Closing a background-check loophole nobody meant to create
The committee opened with H.R. 3100, the CHILD Act, sponsored by Rep. Fry (R-SC). It fixes a genuine oversight: a 2018 law meant to update the National Child Protection Act accidentally redefined “provider” as “covered individual” in a way that excluded contractors (bus drivers, tutors, contracted nurses and therapists) from background-check eligibility. Services like HopSkipDrive, which arranges vetted transportation for children who can’t use a normal school bus, can request checks on employees but not on the contractors who actually do the driving.
Ranking Member Jamie Raskin gave the bill an unreserved endorsement, calling it “a common sense bipartisan fix to close a gap in federal law that puts vulnerable populations like children at risk.” It passed 23-0.
Low stakes, real comity: the fossil bill
H.R. 9191, the National Fossil Act, would designate the American mastodon as the national fossil mammal and the T. rex as the national fossil dinosaur. It’s a low-stakes bill, and the floor treated it that way, in a good sense. Raskin took the bait on the “extinct relative of the modern elephant” line, joking that the mastodon’s modern political variety “faces extinction itself.” He then noted, straight-faced, that recent research suggests the T. rex’s ancestors crossed a land bridge from Asia: “establishing that immigrants were making America great even before humans walked the Earth.” Members traded state-fossil trivia (Maryland’s Astrodon, Vermont’s mammoth and whale fossils, California’s La Brea Tar Pits) for several minutes.
The bill passed 22-1. The lone “no” came from a Republican member who, while conceding “it’s all maybe done in good fun,” said he couldn’t justify floor time on a symbolic bill and voted no as self-described “a crusty fossil.” It was a joke delivered with a real recorded vote behind it. The comedy didn’t cancel out the consequence.
The most substantively detailed bill of the day
H.R. 7834, the Safe Cloud Storage Act, sponsored by Rep. Lee (R-FL) with Democratic co-leads, extends liability protection to vetted third-party cloud vendors who store child sexual abuse material (CSAM) as evidence for law enforcement. Investigators are currently stuck using thumb drives and local servers because vendors won’t accept the legal risk of touching this material, even when they’re doing exactly what the law enforcement agency asked them to do. The bill requires NIST-consistent cybersecurity standards and a 30-day DOJ notification when a vendor signs a new law-enforcement contract, while explicitly withholding protection in cases of vendor misconduct or negligence.
Multiple members cited NCMEC’s CyberTipline volume figures (reports grew from roughly half a million in 2013 to nearly 30 million in 2021) as the scale of the problem the bill addresses. It passed 24-0.
A small, unglamorous fix with a concrete number behind it
H.R. 625, the Local Access to Courts Act, authorizes federal court sessions in College Station, Texas and El Centro, California, communities more than 100 miles from their current courthouses in Houston and San Diego. It costs nothing extra because the federal facilities already exist. Raskin noted it would help “over 13,000,000 Americans gain equal access to justice,” a figure sourced to his own floor statement, not an independent count, but the underlying logic (long drives for jury duty and court access) isn’t in dispute. It passed 23-0.
The Arctic Frost story: the day’s real news
The most legally consequential bill of the day was H.R. 9098, the Congressional Records Protection Act. It grew directly out of a story that checks out.
Chairman Jordan opened by describing a DOJ letter the committee had received the day before the markup: Special Counsel Jack Smith’s investigative team had obtained not just phone metadata but the actual text-message content of 44 members of Congress (Republicans and Democrats) and had done so by bypassing DOJ’s own filter-team process, the internal safeguard meant to screen out privileged material before prosecutors see it. Not metadata — actual content. That distinction is what makes this a story worth telling precisely, rather than a vague “they spied on Congress” claim that collapses under scrutiny.
This is independently confirmed, not just a committee talking point. Sen. Chuck Grassley’s Senate Judiciary Committee released DOJ records on July 14, 2026, the day before this markup, confirming that Smith’s team “apparently bypassed the Filter Team and directly accessed” text messages involving 44 members of Congress. The messages came from a June 2023 subpoena to the National Archives (NARA), not a subpoena to the members themselves, and covered White House personnel communications from October 2020 through January 2021. One important piece of context the committee didn’t dwell on: the 44 members include Grassley and Johnson themselves, along with Cornyn, Graham, Hawley, McCarthy, and Jordan. This is a story told by people with a stake in it, not disinterested investigators. That doesn’t make it false, but it does make it worth reading with that in mind.
The bill itself would require judicial probable-cause certification before DOJ can seek a member’s or staffer’s communications, mandate notification to the member, and impose a 30-day delay before investigators can review anything they collect. Raskin negotiated a bipartisan amendment adding a special-master review process, so the bill doesn’t accidentally block legitimate investigations that have nothing to do with a member’s official legislative work. His example was an FBI investigation into foreign agents cultivating a relationship with a member for espionage purposes.
Not everyone thought the bill went far enough. A Kentucky Republican, self-identified as one of the 44 whose texts were obtained, argued the bill should require warrants rather than subpoenas and objected to it implicitly conceding the “third-party doctrine”: the legal theory that records held by a third party (like a phone carrier or, here, the National Archives) don’t carry the same privacy protection as records in your own possession. A New York Democrat pushed back hard on the framing, arguing that metadata collection “is not spying” and defending Smith’s investigation as legitimate, standard corroborating evidence-gathering in a case where Smith had “far more than probable cause.” The two sides didn’t resolve that disagreement, but they did agree on the bill’s text.
That’s a different kind of disagreement than the ones in the next two items. Nobody in the room disputed that Smith’s team accessed the texts, or how: a subpoena to the National Archives, not a warrant to individual members. The fight was over how to characterize an agreed-upon set of facts: spying versus standard investigative protocol. The sanctuary-jurisdiction and Chloe Cole Act debates that follow have the opposite problem. There, the underlying facts themselves are contested, not just the label applied to them. H.R. 9098 passed 22-1.
The Contested Half: Where 70% of the Runtime Went
The next two items were different in kind, not just degree. Both were framed as being about specific policy questions (sanctuary-jurisdiction prosecution and gender-affirming care for minors), but the floor debate on both spent much of its time on tangents: immigration enforcement deaths, healthcare premiums, the Iran war’s cost, congressional approval ratings. That’s a normal committee-markup pattern (members use their five minutes for broader messaging), but it means a lot of what got said in the room needs a closer look before it gets repeated as settled fact.
H.Res. 1394: sanctuary-jurisdiction prosecutors
This non-binding resolution condemns state and local prosecutors who factor a defendant’s immigration status into charging decisions. The resolution’s sponsor, a California Republican, built his case around two named victims: Stephanie Minter, murdered in Fairfax County, Virginia by a man with dozens of prior arrests who’d repeatedly been released rather than turned over to ICE, and Sheridan Gorman, an 18-year-old college freshman shot in a Chicago park by a man who’d previously been released after an earlier arrest under the city’s sanctuary policy. He and other Republican members named specific prosecutors (Fairfax’s Steve Descano, Philadelphia’s Larry Krasner, the Boulder County and Queens DA’s offices) and quoted their own public statements about weighing immigration consequences in charging decisions.
Raskin and other Democratic members opposed the resolution as toothless political messaging with roughly three weeks before the midterm recess, and redirected to two recent ICE-involved deaths: Lorenzo Salgado Araujo, killed July 7 in a Houston traffic stop, and Joan Sebastián Guerrero, killed July 13 (two days before this markup) in Biddeford, Maine. Both killings are real and independently reported, and in both cases DHS itself said the men killed were not the actual targets of the enforcement operation. For Salgado Araujo, that confirmation traces to Rep. Sylvia Garcia (D-TX), who said DHS told her directly that he wasn’t the intended target; contemporaneous reporting on Guerrero’s death in Maine cites similar DHS statements to local press.
Where the debate needs correction: a New York Democrat cited a Wall Street Journal statistic that “92 percent of detainees… had no criminal convictions,” repeated later by a Vermont Democrat. That figure is real, but it’s being used to describe the wrong population. The 92% figure describes the growth in ICE detention during a specific roughly 3.5-month window (September 2025 to January 2026): of 11,296 new detentions in that window, only 902 involved someone with a criminal conviction. It is not the share of the total detained population. The more relevant number for a general claim about who’s currently in ICE custody is 70.6%, per TRAC’s tracking as of July 11, 2026 (46,436 of 65,765 detainees with no criminal conviction), still a clear majority, but a meaningfully different figure than the one used in the room. The gap isn’t evidence either number was invented. It’s what happens when a snapshot of new arrivals to a system gets treated as representative of the system’s entire population. A short-window growth figure can skew toward whatever category of enforcement activity happened to spike in that window; the steady-state total reflects a much longer accumulation. Citing the narrower number without saying so isn’t lying, but it is, functionally, picking the more dramatic of two real statistics.
The DHS/ICE budget also got no consistent number: Raskin cited $240 billion, a Washington Democrat cited $210 billion then later $170 billion in the same debate, another member cited $250 billion. None of these is simply wrong (it depends what’s being totaled), but none was sourced in the room either. DHS’s own July 2026 statement puts the OBBBA-specific border/immigration allocation at $165 billion; outside trackers like the Brennan Center put combined ICE-plus-CBP spending, including pre-existing appropriations, at more than $200 billion. If you’re going to cite one number, those are the two defensible ones.
One more figure needs correcting from this same stretch of tangents: Raskin cited the Iran war’s cost at “$135 billion,” and separately, later in the day during the Chloe Cole Act debate, at “$132 billion.” Neither matches independent reporting. Defense Secretary Pete Hegseth’s July 21 congressional testimony put the direct cost at $37.5 billion; CSIS’s independent estimate as of late June was $34-42 billion. Raskin’s figures run three to four times higher than either: they may be folding in broader Middle East posture costs, or conflating the war-fighting cost with the separate $68 billion Americans have paid in higher gas and diesel prices since the war began, but as stated in the room, they don’t match a named, independent estimate.
Raskin also cited Congress’s approval rating at “86 percent” disapproval. That one holds up. Gallup’s April 2026 poll put congressional disapproval at 86%, tying the all-time record. H.Res. 1394 passed 14-9, the closest vote of the day.
H.R. 7651: the Chloe Cole Act
The final and longest item, running more than an hour, was the Chloe Cole Act, sponsored by Rep. Onder (R-MO). It would create a federal civil right of action letting minors and their parents sue doctors, clinics, and hospitals for providing “covered interventions” (puberty blockers, cross-sex hormones, and surgery, explicitly not counseling or therapy) as part of gender-affirming care. It’s named for Chloe Cole, who testified before the committee in 2023 about her own detransition after receiving these treatments as a young teenager.
Both sides came in with citation-heavy arguments, and both need correction before repeating anything from this debate as established fact.
Onder’s central citation was a study he described as “Ruska et al., Acta Paediatrica, July 2026,” which he said showed gender-affirming procedures make “serious mental disorders get dramatically worse.” The study is real, but the details are off in ways that matter. It’s Ruuska et al., published April 4, 2026, not July. It’s a study of psychiatric morbidity among adolescents referred to gender identity services in Finland, compared to the general population. It found meaningfully higher psychiatric morbidity in the referred group both before and after referral. What it does not establish is that treatment caused worse outcomes: the study doesn’t disentangle the referred population’s pre-existing conditions from any effect of treatment, and a formal Letter of Concern was published in the same journal raising exactly this methodological objection.
Onder also cited the UK’s December 2024 indefinite ban on new puberty-blocker prescriptions (tied to the Cass Review), a March 2023 Norwegian Healthcare Investigation Board report, and a February 2026 American Society of Plastic Surgeons statement recommending delaying related surgery until age 19. Those citations check out as described.
Raskin countered that “all major medical associations agree that evidence based gender affirming care is necessary and in some instances, life saving,” cited a Harvard T.H. Chan School of Public Health study finding that US surgeons largely aren’t performing gender-affirming surgery on minors, and referenced a 9% detransition figure from what he called “the largest-ever survey of transgender Americans” (almost certainly the 2022 US Trans Survey, though it wasn’t named on the record). He also described a $2 million New York jury verdict for an unconsented double mastectomy performed as part of gender-affirming care. That case is distinct from Chloe Cole’s own pending lawsuit, which the committee noted has a trial date set for next year.
The sharpest exchange involved suicide statistics. A Texas Democrat cited a figure that suicide attempts among trans and non-binary youth rose “by up to 72 percent” in states that passed anti-transgender laws: a real, peer-reviewed 2024 Trevor Project finding. Onder submitted a rebuttal from the Society for Evidence-Based Gender Medicine, arguing the study lumps different categories of law (sports restrictions, ID-document rules, medical-transition bans) under one “anti-transgender” label, letting a finding driven by one law type stand in as evidence about all of them. Both the study and the critique are real. The Trevor Project study discloses its authors are current or former Trevor Project employees, a genuine conflict-of-interest disclosure, not an attack line. Neither side’s framing is settled by the other’s citation.
Both parties also submitted dueling documents accepted into the record without independent authentication in the room: Republicans submitted the Finnish study, a plastic surgeons’ statement, and a National Review article; Democrats submitted a 170-signatory civil rights letter, a 115-organization opposition letter, an HRC statement, and statements from 30-plus medical associations. H.R. 7651 passed 15-8.
The Stats That Didn’t Hold Up
A quick reference, since these are the numbers most likely to resurface elsewhere without the correction attached:
- ICE detainees with no criminal conviction: cited in the room as 92%. That’s the growth-window figure for a specific 3.5-month stretch. The total detained population, per TRAC as of July 2026, is 70.6%.
- DHS/ICE budget: cited as $170B, $210B, $240B, and $250B by four different members in the same debate. DHS’s own OBBBA-specific figure is $165B; combined ICE-plus-CBP spending is estimated above $200B by outside trackers.
- Iran war cost: cited as $132-135B by Raskin. Independent estimates (CSIS, and Hegseth’s own July 21 testimony) put the direct cost at $34-42B.
- The Finnish gender-clinic study: cited as “Ruska… July 2026.” It’s Ruuska et al., published April 2026, and it measures psychiatric morbidity in a referred population. It doesn’t establish that treatment itself causes worse outcomes.
What This Markup Actually Tells You
Strip away the framing on both sides and here’s the shape of the day: five bills moved through committee with real, substantive bipartisan agreement, including one (the Congressional Records Protection Act) that grew out of a genuinely surprising, independently confirmed story about how the executive branch handled Congress’s own communications. Two items then consumed most of the session’s time because they touch questions where both parties have settled on numbers and framings that outrun what the underlying sources actually say. That’s what happens when five minutes of floor time doubles as a campaign message with the midterms weeks away.
None of the seven bills’ actual text was read into the record (all were “considered as read” by unanimous consent), so anything characterizing what a bill “says” should be checked against the text on Congress.gov, not a sponsor’s or opponent’s summary.
If you want to know what your representatives are actually voting on and arguing about, not the six-second clip, not the press release, that means watching the whole hearing or reading a transcript like this one. Citizen Portal exists to make that easier: we track committee markups as they happen and separate what was said from what holds up. Follow your delegation’s votes and see the source material yourself at Citizen Portal.
Leave a Reply