Tiboko-Tifuh: She Got Her Bond Hearing. The Court Said It Did Not Count.

A federal judge orders a bond hearing and tells the government it must prove you are a danger or a flight risk. The hearing happens on time. The immigration judge denies bond anyway. On paper, the order was followed.

Larissa Tiboko-Tifuh had been in custody for more than nine months when that happened to her. The government brought two documents to her hearing. She brought 36 pages. The immigration judge kept her locked up, reasoning that she would have nothing to “come back for.” Then the federal judge ordered her released and explained how a hearing can take place “perhaps in form, but not in substance.”

The case is Tiboko-Tifuh v. Noem, No. 3:26-cv-01215-JO-DEB (S.D. Cal. June 4, 2026), ECF 15, an order by District Judge Jinsook Ohta. It answers three questions that matter after a bond denial. Can a federal court review an immigration judge’s flight-risk finding? What does the government have to show when it carries the burden? And must a person in custody appeal to the Board of Immigration Appeals first?

Nine months in custody, and a federal order

Tiboko-Tifuh is a citizen of Cameroon. She entered the United States without permission on June 5, 2025, and immigration officers detained her immediately. They found she had a credible fear of persecution if sent back. The Department of Homeland Security moved her into regular removal proceedings and kept her in custody, as the court’s March 12, 2026 minute order recounts.

On January 28, 2026, an immigration judge ordered her removed to Uganda. She appealed to the Board of Immigration Appeals, or BIA, on February 21. Four days later she filed a habeas petition, a request that a federal judge rule on the legality of her detention.

She was held under 8 U.S.C. section 1225(b)(1), a part of the immigration statute that provides no bond hearing for people who come through this credible-fear screening. The court granted her petition anyway, on constitutional grounds. It looked at how long she had been held, how much longer she might be held and whether she had caused delay in bad faith. She had been held for more than nine months. Her appeal could add several more. No one had shown bad-faith delay. Detention that long without a bond hearing, the court wrote, “gravely risks the erroneous deprivation of Petitioner’s liberty interest,” especially when the record contained no evidence that she was dangerous, that she would run or that the government had another legitimate reason to keep her. The court held that her detention without a bond hearing violated the Fifth Amendment’s Due Process Clause.

The order put the burden on the government

The court gave the government seven days and set the rules. Tiboko-Tifuh had to be released unless the government proved, by clear and convincing evidence, that she posed “a danger to the community or a flight risk that cannot reasonably be addressed by bond or conditional release.” If no compliant hearing took place by March 19, she was to go free.

Those terms reversed the usual arrangement. In an ordinary immigration bond hearing under 8 U.S.C. section 1226(a), the detained person must convince the immigration judge that release is warranted, as the Board explained in Matter of Guerra, 24 I&N Dec. 37 (BIA 2006). The Ninth Circuit upheld that arrangement in Rodriguez Diaz v. Garland, 53 F.4th 1189 (9th Cir. 2022), and rejected a section 1226(a) detainee’s demand for a second hearing with the burden on the government. It stressed what section 1226(a) offers from the start, including a bond hearing and a right to appeal it. Tiboko-Tifuh had spent nine months with no bond hearing at all. Here, the federal court’s order made the government do the proving.

Two documents against 36 pages

The hearing took place on March 17, two days before the deadline. The government offered two documents: a Notice to Appear, the charging paper that starts removal proceedings, and a Form I-213, the record officers prepare about a person they apprehend. Those papers noted that she had entered without inspection and been apprehended. They said nothing else about whether she would show up for court.

Tiboko-Tifuh’s 36-page bond packet said a great deal more. It gave a verified address where she would live and could be reached. It showed nine months in immigration detention with no criminal conduct and no noncompliance. It identified a financially secure sponsor committed to supporting her attendance at future proceedings. And her statement told the immigration judge that she had fled persecution and torture by the Cameroonian military, that colleagues had been killed and that she had been kidnapped. The court read those facts as further evidence of her personal incentive to stay in the United States and pursue her asylum claim.

Nothing to “come back for”

The immigration judge found her a flight risk and kept her in custody. According to the court, he did not engage with the evidence from either side and did not explain how the government’s proof outweighed hers. He reasoned instead that she was unlikely to win her appeal and would therefore have nothing to “come back for.”

The court called that speculation, and an “unsupported inference” that could not take the place of evidence.

How did the federal judge know what happened in the hearing room? Tiboko-Tifuh filed a declaration from Vilena Ramini describing it. The government did not oppose those allegations, did not challenge the declarant’s credibility and did not submit a transcript. So the court took the declaration as its account of what occurred.

The federal court could review the flight-risk finding

The first hurdle was power. Congress has barred federal courts from reviewing decisions that are purely discretionary under the immigration statute, in 8 U.S.C. section 1252(a)(2)(B)(ii). A bond decision looks like one of those. The court held that the bar did not stop it here, for two connected reasons.

A court keeps power to enforce its own orders, including an order granting habeas relief on conditions. And the Ninth Circuit has already said an immigration judge’s dangerousness finding is a reviewable “mixed question of law and fact,” meaning a legal standard applied to a set of facts, in Martinez v. Clark, 124 F.4th 775 (9th Cir. 2024). The Board’s nine factors guide that finding. That standard is “malleable and involves agency discretion,” the court acknowledged, but it is still a legal framework that a reviewing court can check. Flight risk, the court reasoned, works the same way, because the same factors inform it. So the question it could answer was narrow and concrete: Did the immigration judge hold the government to the standard the court had ordered?

The narrowness matters. Martinez does not let a federal judge reweigh the evidence, and it left a dangerousness finding in place. In March, another judge of the same district turned away a detained asylum applicant’s attack on his bond denial. The court said it could not “reweigh [the] evidence,” and it dismissed the claim because his BIA appeal was still pending. Tiboko-Tifuh made a different complaint. She did not ask the federal judge to decide that she looked safer than the immigration judge thought. She said the immigration judge had never put the government to its proof. Martinez leaves room for that. An agency need not discuss every piece of evidence, the Ninth Circuit said, but it gets no benefit of the doubt when it “fail[s] to mention highly probative or potentially dispositive evidence.”

Why two documents could not carry the burden

Clear and convincing evidence means more than a hunch and more than a tie. To carry it, the government had to show it was “highly probable” that she would miss future proceedings. The Supreme Court’s image for that standard, which the order borrows from Colorado v. New Mexico, 467 U.S. 310 (1984), is evidence that instantly tilts the “evidentiary scales.” An immigration judge weighing that question looks at factors like these: a fixed address, length of residence, family ties, work history, record of appearing in court, criminal record, immigration violations, attempts to flee and manner of entry. When the proof falls short, the government loses.

The government’s only flight-risk fact was how she came in. Even assuming the immigration judge relied on it, which the record did not show, the court said it could not do the job. “Entry without inspection is common among asylum seekers,” the court wrote. Without individualized evidence about her entry or some past failure to comply, it did not support an inference that she would fail to appear. The court drew on Singh v. Holder, 638 F.3d 1196 (9th Cir. 2011), where the Ninth Circuit said a fact “common to all detainees” could not by itself be clear and convincing evidence of flight risk.

Singh is closer than the order’s short citation shows. The fact common to all detainees there was a final removal order, which the Board said diminished Singh’s incentive to appear. The Ninth Circuit called that relevant but not enough. That is very nearly the immigration judge’s reasoning here: She would probably lose her appeal, so she would have no reason to come back. The Ninth Circuit has since said, in Rodriguez Diaz, that Singh rested on an assumption the Supreme Court later rejected, that the statute itself implied these bond hearings, and it left open which parts of Singh remain good law. The June order described the burden as both “ordered by this Court and required under Ninth Circuit law.” As we read it, enforcement did not turn on the Singh question, because the burden was already a term of the court’s own order.

The court went further. Even if that thin showing could support an inference, her evidence outweighed it. And when the government wins with almost nothing, and the immigration judge offers no reason for setting aside substantial evidence on the other side, “only one conclusion follows.” The judge had not put the burden on the government at all. For that step the court cited Ninth Circuit cases it described as finding legal error in an unexplained decision, which is different from a disagreement over weight. The June order cited Yepes-Prado v. INS, 10 F.3d 1363 (9th Cir. 1993), as an example of legal error: The immigration judge had “failed to offer a reasoned explanation” for treating one adverse factor as outweighing all the favorable evidence.

Why she did not have to go to the BIA first

Ordinarily there is another stop, and the hardest case for Tiboko-Tifuh was Leonardo v. Crawford, 646 F.3d 1157 (9th Cir. 2011). The facts are close. A federal court ordered a bond hearing or release. The immigration judge denied bond. Leonardo went back to the federal court and argued that the hearing had not conformed to the order. The Ninth Circuit called that a “short cut” and said the proper path was an appeal to the BIA, then federal court. It also said the district judge had no obligation to take up his challenges as a question of compliance, because that judge had found that the government “complied exactly” by holding a hearing.

The June order cites Leonardo for the court’s power to review compliance and analyzes exhaustion under other cases. As we read the two decisions, the difference lies in what the first order said. The order in Leonardo, as the Ninth Circuit describes it, called for a hearing. Tiboko-Tifuh’s order named the burden and the standard, so her objection was no new argument. It was the order’s own term. The government still argued that she had to go through the BIA first, a requirement called exhaustion.

The court declined to require it. Exhaustion in this setting is a rule courts apply as a matter of judgment, not a statutory command, and they look at whether the agency’s expertise is needed, whether skipping the agency would encourage people to bypass it and whether the agency is likely to fix its own mistake. The court doubted the first and the last.

Federal judges decide flight risk and danger all the time, the court said, particularly in the Southern District of California, where the court regularly decides detention and bail for noncitizens. And it had “serious doubts” that the BIA would correct the error. Citing another judge of the district, whose order drew on NPR reporting, the court said the BIA had been “upholding 97% of all immigration appeals in 2025, including bond determinations,” and said it could not conclude the Board would do more than “rubberstamp IJ bond denials.” That earlier order, Perez Velasquez v. Bondi, No. 26-CV-01759-GPC-DDL (S.D. Cal. Apr. 16, 2026), ECF 11, described the reporting as finding that the BIA agreed with the Department of Homeland Security in 97 percent of cases. It also said such evidence, on its own, was not enough to waive exhaustion, while finding that the statistics supported the view that the BIA was far less likely to give relief.

Even if exhaustion were proper, the court said it would waive it. BIA review would likely mean weeks or months more of detention the court had already found unconstitutional, and that was irreparable injury. The order collected figures from other courts, including an “average processing time of 204 days” for bond appeals in 2024.

This part of the order depends on its setting. In Rodriguez-Gonzalez, a California federal court required a BIA appeal where an immigration judge had written a lengthy decision on danger and the petitioner brought a new case without appealing. Tiboko-Tifuh came back to the judge who had set the rules, and the immigration judge had explained almost nothing.

The remedy was already in the first order

The first order said what would happen if no compliant hearing took place by March 19. None had. So the remedy was release, not a second try. The court rested that on its first order and on habeas itself, whose “traditional function,” in the Supreme Court’s words, “is to secure release from illegal custody.” Tiboko-Tifuh moved to enforce the order on March 24. The court heard argument on April 2 and ordered her release that day. The government filed a notice of compliance on April 6 reporting her release, and the court’s written reasons followed on June 4.

The release order resolved the custody issue. Her removal appeal was a separate proceeding.

What this order reaches

This is one district court’s ruling, and it turns on the words of its own earlier order: a court-imposed burden, a court-imposed standard and a release clause. Our analysis of Tzompa Garcia covers a similar enforcement order in Colorado, where the evidence was a list of arrests.

The bigger question is who can get such a hearing in the first place. In September, the Ninth Circuit held in Padilla that a class of asylum seekers held under the same statute after positive credible-fear findings can pursue a due process challenge to the lack of bond hearings. What protections the Constitution requires remained for later. Padilla also recognized that the Ninth Circuit has treated people “apprehended immediately after entering the United States” differently for some purposes. Someone in that position, as Tiboko-Tifuh was described in March, should expect the government to raise it.

What to gather after a bond denial like this one

• The federal court’s exact words. Find the burden, the standard, the deadline and what happens if the hearing falls short. Tiboko-Tifuh’s release followed from those terms.
• The government’s exhibits. List each one and write down what it says about appearing in court. If the answer is nothing, that is the argument.
• Your own bond packet, as filed. Match it to the factors: address, sponsor, family, record of compliance and the reason you have to appear.
• The immigration judge’s actual reason. Get the custody order and any written memorandum, and note what evidence the decision does not mention.
• Proof of what was said. Request the recording or transcript. Where none is available, a sworn account from someone who was there may matter. This court relied on one the government never rebutted.
• The clock. A bond appeal to the BIA has its own filing deadline. Whether to appeal, move to enforce or both is a decision to make right away, with the order in hand.

For representation in immigration custody and removal cases, see our deportation defense practice.