Oral argument — United States v. Yi-Chi Shih

Cal Shoebicki Cal Shoebicki, on behalf of the United States, good afternoon. May it please the Court, that is indeed my plan. I plan to try and split my time, address the government appeal first, and then reserve about half my time to respond to the defense appeal and the arguments that are made to the Court. With respect to the government appeal, the regulations here have a stated purpose, which is protecting the United States from threats to its national interest caused by the export of items, items like high-powered, military-grade radar chips. Despite that, the District Court here adopted a definition of the word, rated, that it admitted, quote, was not in harmony with the purpose of the EAR, and would allow, quote, an informed person who is purposefully seeking to evade the EAR to do so by simply deferring testing of the item until after it is exported, close quote. So let me pose to you the classic textualist argument here. Justice Kagan, I believe. Rated or designed? The text of the specific EAR here says rated to operate. Rated or designed appears in other parts. Rated or designed to operate. Rated or designed is in a different part of the EAR. But that's what the District Court relied on. Yeah, it says rated or designed, correctly? Correct? Not the ECCN that is at issue here, but other parts of the EAR use rated or designed. Right, and this one says rated. That's correct. And the judge says, because other parts of the EAR use rated or designed, rated must mean something different than designed under the canon, applying the venerable canon on surplusage. Why was he wrong? Your Honor, he was wrong for several reasons. I agree that that was a strong motivation for the District Court. And the judge says, I understand in the absence of the or designed and another later on, I'd read rated the way the government does. But it must mean something different than designed, because we all know that people don't duplicate themselves. This is the problem with textualism, but that's a separate issue. Your Honor, you point out something, a good point there. First of all, that the rule against surplusage isn't some sort of statutory suicide pact that requires blowing up the regulatory scheme like the judge did here, simply because there may be slight, if those two words meant the same thing, then there might be some slight redundancy. The government's position, first of all, is that rated and designed do not mean the same thing, and that's supported both by every single dictionary definition that was put before the District Court. Well, what is the difference between what the government was advocating the meaning of rated meant and then designed? I mean, do you agree that there was surplusage if you adopt the District Court's argument, or do you think there's a way to navigate that to say the rated or designed, those actually have separate meanings in that separate provision? The government's position on this is twofold. First of all, that those two words do not mean the same thing, and that was argued extensively before the District Court. Tell me what the difference is. The difference here, it does seem like it was. You're using rated in the sense of design for purposes. I mean, that's how you... So, if I may, Your Honor, the technology here is, and this is laid out in the papers, but first of all, there's an initial design process where you come up with the design for what this circuit's going to look like. It's like a blueprint, okay? And then you go through an iterative process. In this case, the evidence showed it went over months and months of back and forth on tuning this thing to get to the point where it was sufficiently specified, and this is consistent with the testimony of Carlos Monroy, who is the BIS expert. So your position is that rated can involve something other than post-manufacture testing? Oh, absolutely, Your Honor. Which I think was the District Court's position that rated must mean post-manufacture testing. Well, Your Honor, that's part of... There's sort of two things that the District Court did here. First of all, the District Court interpreted the word rating in a way that was not consistent with the plain meaning, and also not consistent with the fact that you can, through the process of putting together the designs and specifications for a radar chip, know to a high degree of certainty how it's going to operate before it's ever manufactured. It made an error there. The other error that it made was importing a time requirement into the definition, this time of before it was exported. There's nothing in the EAR that says anything about timing as to rating. So there's two errors there, Your Honor. And I want to ask you a separate question on this. So finish what you wanted to say. Oh. So with respect to that, the government's view here is that, first of all, those two words, designed and rated, mean different things. That rating is a more final, formalized specification, which is what the... We have witness testimony at trial that said this. Again, not even conceding that witness testimony is necessary because of the plain meaning here. But that through the process, and especially here on the facts here, there was an extensive period of testing simulation that happened before manufacture, that established that these things would perform in the way that they were intended to. What if they do mean the same thing? If they mean the same thing, then I think that you don't apply the rule of surplusage to give rated this different weird name? See, that's what... I don't want to ask your friend about this, but it does seem to me that the rule against surplusage is not a rule against verbiosity. We often say the same thing twice when we want to emphasize something. And isn't it entirely possible here that rated and designed mean exactly the same thing? Your Honor... Do I have to buy your argument that they mean something different? You don't have to buy the government's argument that they mean different things in order to find for the government here. They could mean the same thing, and this statute or this regulation could still have been satisfied in this case. Because of the exception to the rule of surplusage? Yes. Right. So I thought the government did raise an argument that rated could mean, and in some contexts does mean designed, but that's not the only meaning. And doesn't that support your argument that there really isn't surplusage here? Because when it used... I mean, it's trying to clarify that it's not just designed, but you also said specified to operate at. Wasn't that the other definition that you gave? Yes, Your Honor. And that was the definition that was expert witnessed, Carlos Monroy, who came from BIS, who's involved in the rating and evaluation and license determinations on these... So does that get... I mean, does that get around the surplusage by saying, look, clearly rated encompasses some portion of design, but when they... In this other section of the EAR, when they use it, they're also making clear that it's not just designed, but it's also this broader term... And those other parts of the EAR didn't involve these kinds of circuits. So I don't want a hazard to guess or understand what all goes into the process of making... There's one of them involved... So you're taking the argument is there's no surplusage in this part of the EAR? There's definitely no surplusage. There's definitely no surplusage here. I think it was an inapt analogy. I mean, I suppose you could have different sections of the same act that the terms come in different contexts. Is that sort of your argument? Do you have an argument here? Well, that is an argument. Again, that's an argument that's a little bit down the road. The bottom of the line is even if they mean the same thing, the government's view is that it's not invalid. And also that there's... Is the government entitled to any deference here in the Senate? I mean, this is interesting because you've got a regulation that basically serves as the basis for a criminal act, right? Yes, Your Honor. And so does the government get any deference? I mean, not to talk about Chevron deference, but I mean, is there any type of deference that the government gets in a regulation that it's adopted that wasn't adopted by Congress? There are doctrines in that area, but the government is actually not reliant on deference. In a criminal case, it seems to me the statute itself needs to give fair notice, right? Well, the... I mean, or the regulation itself in this case needs to give fair notice. And this happens in a lot of areas. The Clean Air Act has a lot of enabling regulations that also can cause things to be criminal. But again, it does go, at least to that extent, it does go to the notice question. So there's no deference to the government in a... I mean, I think that's right, but I just... It was a question, but it doesn't really... The government's not relying on deference here, Your Honor. We're not going to cut our teeth on that. So before you leave this topic, and I know you want to save time for the other one, if you are right, we don't need to worry about the objects of the conspiracy. You make a separate argument that even if you're wrong, there's still a valid object to the conspiracy. That's correct, Your Honor. And that's one of the ones I'm holding in reserve for the second part, but yes. You can hold on to it. Even if the court doesn't agree with the government here, which emphatically it should, one other point that I want to make here on this before we go off this topic, and I cede the lectern, is that the district court sort of at length talked about how it was giving plain meaning to this word, but then ignored every dictionary definition that it admitted supported the And instead sort of selectively relied on some testimony from experts at trial to come up with this definition that, again, graphs on extra statutory timing requirement. The notion that you could, as the jury found here, design these circuits in China with Chinese accomplices, trick a U.S. company into manufacturing them, then sort of surreptitiously mail them back to China. But that doesn't have anything to do with designed or rated. Well, I mean, that has to do with the sufficiency of the evidence on the other counts, and I tend to agree with your position that there's plenty of evidence, but that doesn't have anything to do with the definition of the term. Well, it does, Your Honor, respectfully. It does have to do with it in that in a circumstance like this, you get an absolutely absurd result when you take what the district court did, which is kind of cobbled together a definition that creates what itself admitted was a playbook for evading U.S. export controls on high-powered radar chips. Okay. Thank you. I'll be back in a bit. Thank you. May it please the Court, James Furtis for appellee and appellant, Dr. Sheehy. And, Your Honor, I want to pick up with where the government just left off. There is no loophole at all in the EAR by distinguishing rated from designed. The EAR absolutely makes that distinction. Well, not the EAR under which your client was convicted. Well, the neighboring. So, yes, the same EAR under which my client was convicted makes that exact distinction. No, not the one in which he was charged. He was charged not under the EAR that has the designed or rated. He was charged under the one that only said rated, right? No, Your Honor, respectfully. So, there's one EAR. I understand there's one EAR. Right. What conduct was your client charged with? So, maybe to answer the Court's question this way, there are ECCNs, which are categories that trigger licenses, that differ depending on what it is that issue. Some of those ECCNs preclude the export without a license of something designed or rated, whether or not it works. Right. But he wasn't charged with violating that ECCN. Right. It's a long way of saying he was only charged with the one that said rated. But they're almost on the same page. Okay. I understand. I understand. But on the same page doesn't mean he was charged with it. That's correct. I think Your Honor's correct that the ECCN at issue has only rated. Let me explore the government's position for a second. Can something be designed to do something but not necessarily rated to do it? No. I'm sorry. Yes. Yes to that question. Can it? Yes. Something can be designed to accomplish a goal, but the thing won't, when tested, actually perform at those specifications. Doesn't that hurt your argument? That helps our argument. Your Honor. But doesn't that mean that designed and rated can have different meanings? Yeah. And that's why they can be used as an or in the other EAR. Yes, that's right. I agree with Judge Nelson. Right. So they do mean different things. So if they mean different things, then why shouldn't we apply what the district judge said was the dictionary definition of rated to your client's conduct? Because the district judge, the dictionary definition issue came up in the post-trial briefing for the first time ever. The entire issue was resolved by witness testimony. We are supposed to interpret a regulation. Witness testimony is of no help to me in interpreting a statute. We do that as a matter of law. Understood. So stop. Let me finish and then you can respond. Yes. And the question is what does the word rated mean? Yes. And the judge says, well, the dictionary definition of rated means your client violated the EIN or whatever. But I'm not going to use the dictionary definition because it would be surplusage. And so my question is if you agree that rated and designed can be two different things and rated is the only thing in the part of this case that refers to your client's case, why don't we apply the dictionary definition of rated? Because the dictionary definition requires testing post-manufacturer. Tell me what? There's only one. Tell me what dictionary says that rated requires. The one that was presented in the post-trial briefing that cites it's the only one dealing with electronic components. All the others are dealing with completely other. Just tell me what dictionary. Yes. And, Your Honor, I will look for that, but there's one. Give me one second, Your Honor. I don't want to eat up too much time. No, and if it's in your brief, you can. It is in our brief. Because I understood that the district judge concluded that if I were to apply dictionary definitions, this would be covered by rated. But I can't apply dictionary definitions because rated and designed mean the same thing, and therefore it must mean something less, something other than designed. So there were four or five different dictionary definitions that the government submitted for its plain meaning argument. One of them required testing. The others did not. The district court made note of the government's point but rejected it because it would be surplusage to say designed or rated if they meant the same thing. But the defense point is that the plain meaning of the word rated means you must test it. And so to be very precise, the plain meaning of rating can't mean designed. So this argument takes out of context a word that the EAR is very deliberate in its use. So there's an intention to use rated differently than designed. And the most important part is the intro to this loophole argument is inconsistent with the express words of the exported administration regulations, which say we don't want to apply it to research. They say expressly that the EAR is designed. I don't know why. I mean I'm not sure that gets you all the way there. Well, so it does in this respect that to go to the definition of rated during a research design process, which is undisputed. I mean the government says things that are – Yeah, but I don't think that that means – I don't think just because they say this doesn't apply to research that that says, oh, it only applies after the product's been developed. So the – I mean there's a lot. There's a massive play in there. Your Honor, one of the problems with this case, with the defense being denied its jury instructions, was that the court took from the jury the whole issue of was this a commodity or technology. The government's – It seems like that was – I don't know why you say that was taken from the jury. The jury was – they were instructed on the difference between commodity – They were not. That's the problem. So we – this was as close to a directed verdict as ever. We were denied our instructions, which we had submitted – But you were denied – Okay, that's a different issue. Yeah. The jury was instructed what it had to find to find something to be a commodity, was it not? No, it was not. No instruction whatsoever on it? There was instruction 29. Yes, and what did it say? So instruction 29, what it said was you have to find that this was rated to perform within this ECCN. What it did not say, what the judge did not tell the jury, is what is technology? What does the word mean? No, we're talking about different things, so I just want to make sure I'm not completely misconstruing the record. Wasn't the jury instructed about what a commodity was? No. It wasn't given examples, but it was instructed about the word commodity, wasn't it? No, Your Honor, it wasn't. It assumed commodity. So the jury had to decide whether something was technology or commodity without – That's right. It was told without – Let's read instruction number 29 at 1 S.E.R. 112 to 113. Okay. I mean, this says, Certain evidence has been presented that items involved in this case were classified as ECCNs in the 3A001 category. In determining whether the category applies, you should consider the following matters. One, the 3A001 category applies to commodities but not to technology. Correct. Commodities are articles, materials, or supplies other than technology or software. Technology is specific information necessary for the development, production, or use of a product. That seems to completely undermine what you just said. No, Your Honor, because the word research doesn't even – What a technology – That's fine, but that's not what you said to us. You said there was no jury instruction defining commodities, and that is not true. See, now, your argument in your brief was that the definition wasn't good enough to let the jury make distinctions. I understand that argument, but you can't stand here and tell us there was no jury instruction about what a commodity was. The judge said, here's what a commodity is. Here's what technology is. It may have been he didn't explain it sufficiently, but you can't argue that he didn't tell them. So when I referenced 29, that was what I was referring to. I would describe that as not instructing. You would hear it as, yes, instructing. The words have – Well, you might describe it that way. Okay. But you can't say the judge didn't instruct the jury. The universe of – Have you made the argument that the definition of commodities or technology here is wrong? Well – Your argument seems to be that it doesn't encompass your research argument. The fundamental research – the threshold question is what it doesn't encompass. So in the appendix, 29 of the appendix has a decision tree, which I think makes this complex argument relatively simple. Before you get to the analysis in 29, there's a threshold question to answer. That requires a definition that was never provided. To jump to step three of that decision tree, which is what 29 does, it skips over instructions that were required to answer the threshold question. Well, let's be clear. Required, that you think were required, that you think were different. But it's not like there was a model jury instruction. There was no model, although we had proposed one that had been submitted. And the district court said that's already encompassed by existing instructions, including 29. So that's why the decision tree illustrates why it was required. Because I use the word required because without that instruction, the jury jumps to a portion of that tree that is down the path without addressing the threshold issue, which is more important. Because you never get to this commodity versus technology analysis. That instruction 29 is an instruction on what you do later in the process if you decide that regulations apply. And what we were denied and what we had been briefing say that they never applied in the first place. You don't do the analysis at all. Just research. Right. But I thought that this was presented. I mean, this was the argument you made. You made this argument to the jury, right? Yes. No, no, not to the jury. I thought you said the judge. I apologize. We argued to the jury, but the government just said what we had argued is irrelevant. Focus on what the judge told you. That's the law. The government gutted the argument. But we couldn't argue it without the judge describing what the law is. And so when I say no instruction. But, I mean, you, just to be clear, you made the argument in a lot of different ways to the jury that he was a researcher, that's all he was doing, and that was one of his main defenses, right? Yes. So your position is just that, as you just pointed out, that that was just gutted because the district court didn't allow the jury to properly consider that argument. By being denied that instruction, which there was ample factual basis for our instruction. It was legally correct. It was literally copied from the regulations. And the court said no literally the night, at 7 o'clock the night before closing argument. After admitting evidence, like Mattis was admitted to rebut the fundamental research defense. So only on literally the eve of closing arguments, so that on the morning of closing, I'm deleting slides, that we were told that the defense is not going to be permitted. We tried to make it. We had made the defense throughout the trial. Again, it seems like that's, I don't think they were saying that the defense wasn't permitted. I think they were saying this is how the defense comes up. I mean, technology. Maybe your argument comes down to there's a difference between research and technology. That's your argument. First of all, that's a part of the argument. I don't believe the jury cared that Dr. Shih was a researcher. Being a researcher under instruction 29 is irrelevant. It just doesn't matter that you're a researcher doing research. Which, by the way, the EAR says in an introduction that it does not want to deter international collaborative research. Dr. Shih published with the alleged co-conspirators. He wasn't charged with international collaborative research. He was charged with shipping items to China. Without a license. Without a license. That wasn't required, is what we wanted to argue. It wasn't required. It's required for a commodity, is it not? Right, but not for technology. Not for technology. And the judge told the jury it's required for a commodity but not for technology. So my question is how were you forbidden from arguing that it wasn't technology? Because the technology has to have a meaning. But the judge defined technology, did he not? So, no. This is the issue. So the word research is. . . Let's be clear. He defined technology. He didn't define it in the way that you wanted it defined. Did you object to instruction 29 and say that definition of technology is wrong? Yes, multiple times. At least four or five times. So, I mean, in the sense that there was all of the briefing on why we were entitled to our instructions, which the judge wouldn't agree with, the issue is clearly preserved. In fact, there's more. . . Just searching the word jury instructions on the docket, I think we filed four different rounds of briefing throughout the trial and didn't get the ruling until 7 p.m. the night before closing. What the jury heard was that jump. . . Effectively, this is argument now. It wasn't in instruction 29. Was that jump to step three of the decision tree, which we've presented to the court in page 29 of the appendix, skip over the whole issue of do the regulations apply, yes or no? Well, but they apply to commodities but not to technology, correct? Yes. Okay. And technology, what is in the definition of technology here given to the jury that prevented you from arguing that the things that were shipped overseas weren't commodities but were just research necessary for the development of production or use of a process? So that the technical data includes, quote, models and engineering designs and specifications. . . But aren't those things necessary for the development, production, or use of a product? Yes. But the fact that it could be manifest in a physical item is something that is important for the jury to know because the witness, Monroy, and the government's argument is based on if you can hold it, it is a commodity. Yeah, but the government's argument aside, it seems to me the judge's instruction allowed you to say, no, this is not a commodity, this is basic research, information necessary for the development, production, or use of a product, and the jury didn't buy it. No, Your Honor, the jury didn't know they were supposed to do the analysis because the government argued the day before closing and won was that their theory of the case is that this is a commodity. They charged it as a commodity. Their theory of the case was that it wasn't a commodity, it was technology. It was charged as technology. That was the big about-face. This was charged as Dr. Xi sending technology to China. We expected this. . . That's an indictment? Yes. Yes, it is. And the indictment charges this as technology and if I can have a moment, I will give you the cite. But the, in fact, give me one second. Am I right that the instruction to the jury was about what I call the 3A001 category? Yes. Is there another category under which you were charged? No. Okay, so the only category under which there were charges is the one where the judge told the jury it covers commodities but not technology. Without further defining them. Understand he didn't say everything else you wanted to say. I'm just trying to get the facts straight. That's correct. Okay, so whatever the indictment said, did you want to be charged with technology too? I wish they would have. They did charge with technology. The jury wasn't instructed with respect to technology that you could be convicted. No, no, no. The technology doesn't require a license at all. I know. Right. I know. The jury wasn't told you can convict this guy if you find that what he was exporting was technology. That's right. They were told you could not convict the jury, you could not convict if you find that what he was exporting was technology, correct? Right, without defining it. Okay, so I'm still back to the question and I'm not sure I've gotten a good answer from you on it. Why isn't this broad definition of technology, information necessary for the development, production, or use of a product, sufficient to cover fundamental research? First of all, it's based on the word information, and it's much more than information. It's much, much more than information. And it's the entire process before something becomes a commodity. This loophole argument, can I address it? Because maybe it will crystallize this point. I must say I'm not worried about the loophole argument. I'm much more worried about the definitions of the word rated and designed. I sort of feel like you wanted to pitch this in the terms you wanted to pitch it. The judge said, no, we're going to pitch it in this term. They cover, they overlap, maybe they're not exact, but the question we've got to grapple with, I think, is the way that the district court framed it in its definition of technology, did that restrict you from making the same argument you're making to us now? And I'm not sure it did. If the court will consider the jury instructions we proposed, which were not – Well, they may have been better, but that's not enough to get a reverse law. The question is whether or not these instructions fairly covered the topic. But what I submit is the instructions that we proposed were required to fairly cover the topic because without the instructions defining the terms technology further, the witness, Monroy, and the government's argument that if you hold it, it requires a license because otherwise these things can end up in China without a license, that was wrong. And just like the temporary export, you actually – there's an exception. To send something to China for testing is – Go ahead. You're in jury instruction at 1 SCR 143. Now, maybe there was another one, but defines technology, and one of the definitions of technology includes – one of them includes a result from fundamental research. But there's all these others. And so that's why I don't know why you still were precluded from making the same argument. I mean, by your own definition, technology includes research and other things. So research and other things, including what the government is arguing, constitutes testing. Like, for example, simulations. But that might be a reason why the district court rejected it. No, the government is arguing that these chips were tested. These were models for testing. They did not even work according to the design plan. Well, but that's why I'm getting confused here, so you've got to help me. I'm sorry. And not argue with me, but just help me. The charge here was that these chips were sent to China. Put aside the other counts. We're just talking about the counts one and two. That the chips were sent to China without complying with the EARs, which is to say you can get an export license or make a filing with respect to them. How were the chips fundamental research as opposed to even your definition of fundamental research deals with information? It doesn't deal with this physical product, does it? Yes, it does. So that's where the word models becomes so critical. Your intuition is exactly what the juror's intuition would be. Without that added instruction that when you make a commodity chip, for chips that become commodities, those are mass production items. Okay, so tell me what support do you have in the language of the regulations, because the Act just incorporates the regulations, for the notion that selling a model that otherwise meets the definition of an MMIC overseas is exempted because it's just a model. So that's where the definition of technological data becomes so critical, which says that the technology includes models. So that's critical. Where's that definition? 15 CFR 772.1. It's the definition of technical data. And the other related request instruction was that technology includes specific information necessary for the development, production, or use of a product. So you can't use something that is not made. If you're talking about chips, you can't use it unless it's made. You can't test it unless it's made. And then the fundamental research definition is the third, which is in a different place. So I'd like to give this. Stick with me for a second on model. How many of these chips were produced? So for the ones requiring the license that the government's theory is, I believe it was 82. This wafer had sections called reticles. So 82 were produced. There were lots of different reticles testing lots of different ideas. The government's view is that one of the reticles triggered a licensing requirement. 82 identical ones were produced. Yes. Not different models. No, no, no. Right. There's 1,012 chips on a wafer. Of that 1,012 potential design ideas, they're partitioned off in reticles, 82 per. And I believe the government zeroed in on one of the reticles that was basically being researched and tested. I understand, but is it your position that multiple models were produced? Well, so the way that wafers are made. I don't know. The wafer's the model. Your answer is that multiple models were produced? Well, so one wafer is a model. The model contains multiple instances of the research. Is it your position that multiple models were produced? So the minimum. Yes or no? I don't know whether it's right or wrong. I just need to know yes or no. The answer is yes because CREES policies, they won't make less than four wafers. So that's the minimum order you can make. So they produced how many models in your view? Well, it would be 4 times 82. 326 models, 328 models? Yes. You see, you think the statute, you think that language is meant to cover what's in effect an industrial production of hundreds of chips and they can all be treated as a model? So can I please address that with three or four sentences? I apologize, but it's such an important question. It's such an important question that captures the difference between technology and commodities. The cost of these four wafers was over $100,000 because they were models. If this were commodities and people would invest the million dollars to do a full-scale run, the cost per chip for commercial use would be fractions of the amount. The whole idea of electrical engineering is you do all the research you can because it's so expensive to construct the model that you do the planning, planning, planning, and then you run a model to see if it works. Usually it doesn't. These did not. But then you produce 328 models and you send them off and chart at whatever price they're sent off at. But they're valueless. To make the model costs $100,000. But, yes, the minimum number for each time you do a test, you destroy the chip. So by definition, you're going to create more of the things when you're testing them because each test destroys it. These are microchips. And so, yes, there's going to be more than one made, of course. In fact, allocating only one reticle to this particular design idea almost proves that it was a research not ready for use. This government argument that it ended up in the nose cones of an F-35 fireplane was inflammatory and impossible. Okay, put aside your characterization. I want to just ask you one more question. Okay. And this is with respect to the part of the charges that said you shipped off something that was worth more than a certain amount without making the necessary filing. Yes. Does your technology and model argument apply to that also? Yes. You do not need to file EIN for technology. If it's not technology. If it's technology. If it's fundamental research. So what I don't know is, is it possible there's a technology that could somehow not pertain to these chips? I don't know. But for this discussion, yes. Technology means no license or EIN. Okay, that's what I understood. Yes. Thank you. Yes. And the point is that the EIR is drafted by experts. Mechanical engineering experts draft this language for computer chip research. That is why this is so important to get the instruction of what technology is, including models, because I think the intuition, the human nature is to think, like the district court did, if I hold it and it worked, in other words, it tested to perform at post-manufacture, a license was required, and that's not true. The EIR expressly says in the intro, we want to encourage this type of collaborative research. The co-defendants were researchers, and we were able to introduce them, but the jury never saw the papers that were written with the charged co-defendants, the theses, the patents that were published with co-defendants. These were visiting professors doing research and our position, because this is a willfulness crime. You have to know the law which we stipulate and intend to violate, which is what the trial dispute was. Dr. Sheen knows what the regulations are, which is why he knows that for research you don't need a license. Okay. Okay. I think, look, it's been helpful. We appreciate your argument. I mean, I think you're well on. And I consumed all my time on this threshold question when I did hope to discuss all the evidence that was excluded at trial. No, I know. We all hope to have more time, but you used the time you got. I understand. Thank you. Actually, we gave you eight more minutes than the time limit. Thank you. Thank you. It would be helpful for me if you could address the last conversation we were having, and so that I understand the context. I take it the government's position is instruction 29 was good enough to cover the topic. But can you address this model issue for us? Absolutely, Your Honor. I think that I can address both that instruction question and the model question and try and unpack this a little bit. The government's position is that the instruction was absolutely a correct statement of the law and sufficient for a defendant to argue. Now, the district court actually really explained this extremely well in its decisions. It's an ER40 and an ER39 with respect to both fundamental research and this question of models and engineering designs. Now, if you'll humor me, I just want to read what the district court said because I think it actually gets directly at what the court's been talking about and a fundamental disagreement with the defendant's view of the law. The fundamental research exception to the licensing requirements of the EAR applies to an item that is classified as technology and not commodity. As a result, to show that a license for EEI filing was not required because an item was within the fundamental research exception, the threshold question is the proper classification of the item as technology or commodity. As explained above, there was no factual support for the argument that the Cree mimics were technology. So what the district court said is, look, all this talk about fundamental research and the many arguments about, you know, what he was doing with this stuff and what people were doing in China, whether it was research or commercial or whatever else, that only applies if we're talking about technology. And the statute is extremely clear with respect to those definitions. And, in fact, the instructions were given. Bear with me for a second. Yes, absolutely. This is where I'm a little bit confused after listening to your friend. You've got a model. God knows what a model is, but it's you make something and it's something that's exactly like what you want to make later. Can that model be technology? So, yes, technology can include a model, and let me explain what that means in context. And this is, in fact, the court's confusion about this issue and the arguments that the defense made about it here, and these are the same arguments the defense made before the district court, shows exactly why the court did not incorporate that word or engineering designs into the jury instructions because ripped apart from context, they are super confusing. Now, let's look at the definition. Technology in the statute. So this is 17 CFR 772.1, and this is the statute at the time. Defines technology as specific information necessary for the development, production, or use of a product. The information takes the form of, quote, technical data or technical assistance. Now, let's talk about technical data. That's also defined. Technical data is defined, it says technical data may take forms such as blueprints, plans, diagrams, models, formula, tables, engineering designs and specifications, manuals and instructions, et cetera. Okay. So that's in the EAR. That's in the EAR. Why, if the EAR says it may take forms of models, then why wouldn't that be appropriate to put into a jury instruction? Well, what the defense asked to do, well, there's a couple reasons. First of all. Put aside whether or not he correctly, the district judge, correctly rejected their proposed instruction. I'm trying to figure out whether this one is sufficient to allow them to argue that what they shipped were actually models and, therefore, technology. So here when we talk about models, models are, again, let's take a step back. It's information. Okay. So let me try and the difference between a commodity is a thing now and technology is information. It's difficult conceptually to visualize. The idea is, look, this is a thing, right? This case that I printed out, it's a thing. Okay. It also can convey information. That doesn't make it not a thing. Okay. So the question here is. A commodity. Right. This is a commodity, but it also conveys information. Okay. A model in the context here, what a model is, when we were talking about making the mimics, they do a blueprint. They have a software blueprint. That's a model. That model is information. But once you make it a thing. So your position that the physical MMIC cannot be a model? Yes. The notion of model, the definition of model that the defense advocated for and argued about is more along the lines of something that you're going to test on and calling that a model rather than a model as is defined here, which is information. If you think about it like a. . . Say here in the EAR. In the EAR. Yes. I'm sorry, Your Honor. So think about a watch, right? Swiss watch. You buy a Swiss watch. It's a thing, clearly a commodity. You could take it apart and study the way it's made. You could see the gears. You could derive information from it. That does not turn the watch into information. It's still a thing. And so that's the point here with these mimics. That's why I was asking a friend, and that's what I want to return to. The MMIC is a commodity. Yes. And it's not turned into a non-commodity merely because it can be replicated later to build the same MMIC somewhere else. Or because it could be. . . Or because information can be derived from it. Correct. Absolutely. Everything in the world has some aspect that is informational to it. That doesn't. . . You know, a car or a watch. . . All these things you think of as a commodity, like they're physical things. And under the. . . Not just you think of. . . These are very specific definitions in the EAR. So you think model in the EAR means, in effect, here's how you design one of these. Right. The information. That is 100 percent consistent with the definition of technology. But you do seem to be arguing that his proposed jury instruction would have actually been misleading. Yes, Your Honor. That's not what the district court found, though. What the district court said was this proposed jury instruction is already encompassed by the other jury instructions. That's why I'm not including it. Well, Your Honor. . . I mean, there's a lot in here, so I'm sure you'll point me to something else. I mean, it's, you know, I need to look at EAR 39, but I think that it. . . I mean, because I'm sitting here thinking. . . I mean, what I sort of posited was, look, they could still have made the same arguments. They just had to use different terminology. But you seem to be saying, no, you couldn't make the same arguments because the arguments that we just heard are flatly inconsistent with how the EAR operates. Well, you don't get to the question of subsets of technology if the thing is a commodity. So the jury instruction saying, if you think this is technology, the ECCN doesn't apply, is A, a correct statement of the law. B, allows them to fully argue, as the defendant did. The court's seen the record. Defendant argued about this all the time. But then everybody, if I understand this correctly, and this is kind of the loophole argument, you just define everything as a model and then send it all over. I mean, yes, if model means what the defense is sort of positing. But this model thing would be very confusing to the jury. If you pull, look, I read you a long list of terms. Those are all terms that are intellectual constructs, blueprints, ideas. And there's the two words model and engineering designs. Those are the only two that were picked out and asked to be grafted onto an instruction. Well, but I understand engineering designs because that's information. Right. But can you have a model that's just information? Yeah, a software model, Your Honor. It's not built. So you can have a model inside a computer. The plans are not a commodity. Once you take those plans and you put them through an industrial process and you make a thing, then you have a commodity. So the model that sits, it's a conception. It's information. So the government's position is that once you turn whatever research you're doing into a thing and send it overseas, it's a commodity subject to the statute. Well, you don't even have to send it overseas. It's a commodity. I understand. Subject to the EARs. Subject to the EARs. I'm sorry. It's a commodity subject to the EARs, even if it's a prototype. Absolutely. Even if the whole purpose is, well, I put this together and I'm going to send it to you because once you get it, you can do research on it and determine how to do it. Absolutely, Your Honor. If I may, I know that I see that my time is out here. I wonder if I may very briefly. We didn't talk about this. I want to address one issue, which is the claim about the misconduct during a rebuttal, if the court's interested in hearing about that. I'm going to defer to you because I think that's fair that you've been accused. Go ahead. Well, one of my former colleagues. But we, America, have been accused, Your Honor. So the point here is that during the defense's closing, the defense actually had a section in the closing presentation called Tactic 2, Distracting with Fear. And as part of that, the defense showed a slide, which is at 6ER1205, which is a slide from a business presentation that was found on defendant's computer, which includes on it a photograph of an F-35 nose cone radar, missile radars, and U.S. warships with sophisticated radar systems. I'm sorry. That slide was admitted into evidence. Again, it's 6ER1205. And there was testimony from a witness even about the F-35 nose cone radar during trial. So when the defense puts this up on the PowerPoint during closing, it was entirely appropriate for the prosecutor to respond to that. And the reference to an F-35 didn't come out of thin air. It came directly out of the slide. You know, here's my problem with both sides in the case. I understand people get heated. I'm not sure it matters to what use these MMICs would be put in China. They could be used in toy dolls, and he would still have violated the regs, under your view, would he not? Your Honor, that's accurate. So both sides. So you bring on somebody to say, well, they're really useful in military operations, and it doesn't matter whether they're really useful in military operations. Your Honor, there was actually no testimony that they were used. Didn't Mattis say that? He did not say they were used by the military. Didn't he say that the people you're selling them to are involved with the military? No, he didn't. He said that he was not allowed to say that, in fact, by the court, Your Honor. But does it matter to what use they're going to be put? Well, yes, there's a couple things that matter, Your Honor. It matters to the argument that these were just sort of this was research or some sort of, like, other thing. These were chips that were being made for customers in China. So the nature of the customers was actually 100% relevant to the defendant's state of mind, his knowledge, and the willfulness here. So, yes, I mean, respectfully, Your Honor, I disagree there. You see, reading this transcript, it seems to me both sides got off on this sort of, you know, you're unpatriotic, no, you're not unpatriotic, you're monitoring fear, and it seems to me it's a very technical issue, which is whether these items are subject to the EARs or not. And they're either technology or commodity. And if they're a commodity, they're subject, even if they're going to be used in Barbie dolls, right? Yes, Your Honor. I agree with that. Okay. No other questions? No, thank you. Thank you to both counsel. And obviously we have your briefs. We know that not everything was addressed here at argument, but I think argument's been very helpful, and we appreciate your preparation and your time today. The case is now submitted, and that concludes arguments for today. I'll rise. This court for this session stands adjourned.