Download the “Outcomes Rocket Podcast_Jason Novak audio file directly.
Outcomes Rocket Podcast_Jason Novak: this mp3 audio file was automatically transcribed by Sonix with the best speech-to-text algorithms. This transcript may contain errors.
Saul Marquez:
Hey, everybody! Saul Marquez with the Outcomes Rocket. I want to welcome you back to the podcast. Today I have a friend and just an extraordinary person and leader in the healthcare legal space. His name is Jason Novak. He is a partner at Norton Rose Fulbright, and his practice was specifically created to focus on advising entities, both large and small, on the various legal issues that could arise with emerging technologies in healthcare, food, and life science industries, and particular target on convergence technologies. That’s the digital health and personalized precision medicine care area that operate in the intersection of multiple industries. Jason has extensive experience in IP and data rights strategy, which is what we’re going to be talking about in this two-part series. And so with that, I just want to kick it over to Jason to talk to us a little bit about him, but also the two-part series that we’re going to be doing here. Jason, welcome.
Jason Novak:
Absolutely. So, yeah, we started this practice, my partner and I, Roger Coin, we were both in-house counsel at Thermo Fisher, and we saw the development of digital health precision medicine from a long-term strategy and a long-term research perspective inside the company, and we saw that this was going to take off right around this time, and so we started our practice about seven years ago geared towards that. What we’re going to talk about over the next couple of sessions is IP strategy for one and data rights strategy for another. What we’ve learned during this process and running this practice is that there’s a little bit of misunderstanding around what the case law around IP strategy, particularly patents and trade secrets, what it actually means in terms of the ability to obtain patents, and then as a consequence has affected substantially the strategies of IP protection across many companies over the last 5 to 6 years. And we’ve discovered that part of this is probably a lack of understanding of some of these technologies because they are convergences, Saul. So you’re going to understand maybe one half or the other half, but understanding how these two blend together and how having a strategy that comprehensively covers the meeting in the middle, it really appears to have been a difficult thing for the industry to get a hold of, and that’s what we’re going to talk about today.
Saul Marquez:
That’s fantastic. So, folks, two parts, one part IP, second part, data rights. Today we’re focused on IP. So Jason, kick us off. Like big case law, how has it affected things? Give us like a level set here.
Jason Novak:
Absolutely. So, in 2012 to ’14, a set of Supreme Court cases came out that really spoke to the ability to protect therapeutics, diagnostic and painting diagnostics, and then software. Fast forward 5 to 6 years, and digital health precision medicine is effectively the convergence of therapeutic drug discovery development, developing companion diagnostics, and using software to do it. So all of that case law has impacted strategy. But the problem is that when all those cases came down from the Supreme Court, books made judgments on what this meant for software patenting, for diagnostics patenting, for therapeutics patenting, and never really revisited the subject. We know, is that typically when a Supreme Court case comes out, the Federal Circuit will help to really put the fine lines and fine touches on what those cases actually mean. What Roger and I did over the last few months is relooked at all the history of these cases going from 2014, about really around 2016 when the Federal Circuit started hearing these cases coming up that were challenging what happened back in 2012 and 2014, all the way up until now. And what we have seen, particularly when the use of AI and machine learning and a lot of these patents, is that the case interpretation and the case law interpretation and the USPTO interpretation of this has really started to align, and there has been real guidance as to what is necessary to protect these inventions, and what it has led to over the last 2 or 3 years are extremely broad patents. Because people back in 2014 and 2017 decided to back away from the table and say, oh, I’m going to panic, I’m not going to file on this, left a lot of green space for others to swoop in who understood the case law and to get very broad claims. And so now what has become, what shouldn’t have been a race is now becoming a race because insights companies that are using data and AI and machine learning to derive insights, whether to be used for drug discovery or for diagnostics, are all going after the same sets of data, are all looking at these data and using AI models to try to derive insights. Which, you end up happening, is a series of markers that are used to define whatever target or insight you’re trying to get to, and multiple companies looking at those, and so now it’s become a race. Many of these companies could have probably gotten out ahead of this if they were dealing with these issues back in the mid-teens and not stepping away from the table and taking a deeper look at what this actually can mean. And what it’s actually led to is a very aligned understanding of where the eligibility for patenting and precision medicine, digital health is, that’s very aligned with the case law and with the USPTO. On top of that, a lot of what we’ve done in the US is moved in a direction closer to Europe. So now your two largest territories in the world for patenting are more aligned than ever, meaning we’ve limited what we’ve been able to do that more and more closely aligns with what Europe is doing. So for the first time, not only do we have clear understanding of where we can patent in digital health precision medicine, we know that we can write an application that will work in Europe as well, and that doesn’t often happen. If you speak to any European attorneys, patent attorneys, they will tell you about the nightmare of US patents coming over to their side and not having the information they need to do their job. We’ve met them halfway now, so it’s a really great place to be from an IP perspective.
Saul Marquez:
Yeah, that’s exciting to know that alignment is there, and at the same time, as a healthcare business owner, a leader of a company that has AI/ML at its core, should they be panicked? How should they be thinking about this and their strategy moving forward?
Jason Novak:
What they need to know is they need to proceed under the understanding as a baseline that protecting the implementation of AI and machine learning towards a actual value and actual insight and actual end game is absolutely patentable. That needs to be your baseline, that is table stakes. That is, you have to know that. See, the problem is you could do IP strategies, start with the understanding that that isn’t patentable.
Saul Marquez:
Got it.
Jason Novak:
And if that is the case, your strategy already, from a foundation is already in the wrong direction. So understanding that it is patentable is step number one. Once you understand it’s eligible for patenting that those cases in 2012 and 2014 don’t get in your way, if you do it the right way, then you need to assess whether or not you need to patent or not at that point. At that point, then it becomes a more of a strategic trade secret versus patent sort of analysis. And I think I want to correct myself right off the top, it’s not trade secret versus patent, it’s together. It’s using them together. And this is one of the other big misses that’s happened back in the teens, as when folks thought we couldn’t patent, then it became a one or the other proposition, either patent or trade secret. What we have learned over time and working with many different companies is we develop a strategy whereby you want to do both. You want to create what I call a hood line, like a hood of a car, and then everything above that hood that everyone can see or is publicly available are things that you want to patent, or else it’s going to go out into the public domain, and somebody else is going to protect it or take it from you. Things that are underneath the hood, it’s a different question. A lot of these machine-learning models are used for drug discovery. You’re not going to make a ton of money as a company on the tool that’s used to derive the targets, you’re going to make your money on the targets. So when it comes to strategy for companies like that, perhaps you may want to protect the targets, obviously, the therapeutic whatever drugs you discover, as opposed to the engine itself, which you could keep under wraps. But each company, it’s an ad hoc analysis. That hood line differs for each company. Certain companies may have to disclose aspects of their discovery platform in order to get adoption in the industry, in which case that headline has now been dropped, and now that platform that you use to discover drugs is now publicly available, now it becomes a whole new game. Every company needs to look and draw their headline in a different place that is most relevant to them themselves. We have clients that disclose almost everything they do on their website because they believe it increases adoption by having that transparency. Well, then you’re putting a lot of IP at risk because if you publish that on the website before patenting it, you’re going to lose rights in certain areas around the world.
Saul Marquez:
Yeah, that’s really good. Look, I think it’ll be worth level setting, like what happened back in the teens that sort of created the confusion. Can we talk about that? Because I think I want to know, and I know probably a lot of people want to know too, listening to this.
Jason Novak:
Yeah, so what happened was, back in 2012, Mayo v. Prometheus was a Supreme Court case that came down dealing with diagnostic protection, and it developed a two-step test for basically determining whether or not what you thought was an invention was eligible for patenting in the first place, not whether or not it already been invented by someone else. No, whether or not you could even patent it. And the reason is that, and this makes sense, if you discover a new snail, that’s a product of nature, you didn’t invent that. The fact that you discovered something that exists in nature, you can’t patent a snail. And so extend that further on, if you discovered an association between a marker and a maybe a cancer, that association arguably exists in nature, you just discovered it. Now, how you treat someone accordingly or how you modify a treatment or whatnot, based upon that understanding of that association, that’s arguably patentable. So that was what that 2012 case came up with. What do you have to do downstream of this association to protect this association? You’re going to have to do something beyond just saying, look, if you identify this versus the other, you got a problem, then you’re blocking the entire market forever using that association, and that’s what those cases got towards. Now, when it initially came, there was this panic of like, oh, diagnostics are dead. There’s no way we can protect diagnostics because they said you can’t protect associations, and then people run for the hills. Fast forward two years, in 2014, Alice … came out, which basically was a, just a pure software case. So now you’ve got diagnostics on one side, two years later, you got software on another side, put them together, you got digital health, or you got precision medicine. And so on the software side, what the Supreme Court said, and I understand the Supreme Court, they’re not patent attorneys, they don’t handle patent cases a lot, so generally speaking, when they come up with holding the Supreme Court that’s related to patent law, its applicability is much more narrow than people first think. The federal circuit are the patent experts, so when they come up with the whole things, they’re usually regarded with a little more heft, or they help refine what the Supreme Court said. So the Supreme Court came out with a new test or a new way of analyzing software patents, and they basically said, effectively, you can take a series of mental steps to get up in the morning. I brush my teeth and put on my clothes, I get in my car, I drive to school, and then put that on a computer and say, that’s a software invention, because all you’re doing is taking a series of mental steps, and you’re putting it on a piece of software or a piece of or on a computer. And they basically said, you’re not allowed to do that anymore. I, as a lawyer knew pre-2014, if I wanted to get a series of mental steps past the patent office, I just put it on a processor, and all of a sudden, it was transformed to some physical structure, and I was allowed to get a patent. That should have never happened, and the Supreme Court and the Federal Circuit, the USPTO, corrected that with Alice. They said, those days are gone. You actually have to talk about how you’re improving the computer, how are you improving the software, improving its speed, its accuracy, its efficiency, something that’s providing value. You can’t just basically take some steps in your head and put on the computer. The computer must allow you to even do those steps in the first place. Perfect example is AI, machine learning, right? Your brain can’t take a series of inputs and pump out a series of markers like that. That’s what it’s built to do. So as a consequence, AI and machine learning inventions are treated like software inventions in that if you’re improving answer, the quality of the answer, the speed of which you’re getting an answer, the way a computer operates, something like that, the how something is working, you can absolutely patent it. And the Federal Circuit, over the course of a few years, really helped to refine that case, because after 2014 came out and everybody’s, oh, I can’t just put this on a computer, all the software patents are dead, and so they all walked away from the table. And to that, in our precision medicine digital health framework, if you’re head of IP is a PhD in molecular biology, and they’re asked to make a software call on a bioinformatics platform, they’re just going to say, well, Alice said, we can’t plant software anymore. We’re done because it’s not their sandbox anyway. It’s the same way if you’re an expert in software, you’re like a double E, and you’re working for a digital health company, and you got to make a call and a diagnostic, that’s not your world. Your world is software, and so this convergence creates a lot of holes in people’s knowledge. And so understanding how they come together has allowed us to understand that the eligibility requirements are, the eligibility standards are not as high as people thought. In the requirements for the USPTO to reject your patent under those rules, under Alice, and under, has gone way up. Like they can’t just say, you’re rejected. I’m sorry, you’re not eligible under Alice, done. No. You have to give a step-by-step explanation as to why you’re not eligible, which just reduced the number of times it’s been rejected because it’s forcing examiners to work harder.
Saul Marquez:
That’s amazing. Yeah, so the opportunity is big, and we might have patents that should be underway that aren’t, as healthcare leaders, is what you’re telling us.
Jason Novak:
And on top of that, you may have made a decision. You know what, I’m not going to patent this because it’s not eligible, but I got to disclose it as a means for adoption. So you effectively put your invention on the market, you’ve disclosed it, which means you can’t protect it in Europe because Europe has an absolute novelty requirement. It means you’ve got to file before you disclose, sell, or offer for sale. So you made that decision to ship a sale. The US gives you about a year buffer from disclosure of filing, it allows for innovation. You got to get it out there, you can’t wait for a patent. Get it out there, we’ll give you a year buffer. But some of those folks may have run past that year, so you may have exclusivity-providing patents that no longer provide you exclusivity because you never filed on that. And in an area where is how many digital health companies are trying to attack every sort of subsector in digital health, exclusivity is the key, whether it be through your contacts in the industry, in terms of customers on customer list, and things like that, or IP. Exclusivity is going to be the means in which investors are going to have confidence that you’re going to be the one that emerges from this battle because you hold the fort when it comes to intellectual property.
Saul Marquez:
Wow, some good insights there, for sure, Jason. And when we first started the discussion, you shared an interesting framework with me. It was this idea of ideation protection enforcement. I’m a big fan of frameworks because they give us a way to think about things and, in this particular case, IP. Talk to us about that.
Jason Novak:
Yeah, where a lot of processes fail, generally, where all processes fails in the beginning. If you never do it right in the beginning, then it doesn’t matter what you do. You can’t catch up when you start making mistakes upfront in the process. So ideation is the first part of IP protection. You have to have an idea that’s protectable before you protect it, and then after you protect it, then you enforce it, so the three-step process to the half-life of the patent. Part of the problem is educating companies to understand what is a potential invention, a patent. And so what happens is companies self-filter during ideation, oh, that’s not patentable, oh, that’s not eligible. And a lot of these aren’t lawyers, they’re just, maybe been advised from an attorney that comes in for a day, or maybe they read up on their own. There’s plenty of blogs talking about eligibility of software patents. There’s thousands of websites you can read and try to get an understanding.
Saul Marquez:
You could ask ChatGPT.
Jason Novak:
You could try that, if you want, yeah, yeah. That’s a whole other question. That’s a separate topic that we can just …
Saul Marquez:
We got to have a part three for that.
Jason Novak:
Yeah, exactly. But from the, from the perspective of ideation, understanding and educating clients to understand what to look for in their own workflows to capture them is step one. Because as an outside attorney, I’m not privy to what’s going on inside the black box. If I were, I could tell them that’s patentable, let’s not make your calls. But as an outsider, you don’t often have that kind of access so you have to rely on education. So educating as to what to look for, given the case law, has really opened up a lot of clients to possible patents that they never even thought of because their inventors never brought it to them as a lawyer. If you’re at a major company like Amgen or something, you have to rely on your principal investigators bringing you inventions, right? You can’t go talking to all of them, they may be all over the world. So educating your inventors, your PIs at a university, or your principal, whoever it is, your head of your labs, or your chief scientist at a major R&D company, things like that, to know what to look for, it opens up their ability to be more open about providing potential inventions to cover. It creates a different culture. And once you do that, then the legal group or the head of R&D, whoever is making the decisions, actually has something to look, at which point they can apply their sort of headline strategy, or we provide it for them or help them determine what that is, and then they can make that call. It’s eligible? Yes. Is it above or below that line? Okay, now make the call. If it’s above that line, do we still want to do it? If it’s below that line, we still may want to do it. There’s still reasons where the 90% of decisions will be made on that line, but there’s other strategic reasons why you wouldn’t. But until you know what’s eligible, until your inventors are funneling everything that you need to consider, you can’t really make that call. You may get one disclosure to file one patent application. You’re going to file it because you have the budget, but it ended up, it may not, it may end up not being that valuable. But if you’ve got 15 sent to you, that one may go right to the bottom of the list because the other ones are more important than your deploying your budget in the more visible way. That’s where ideation becomes so important.
Saul Marquez:
Love that. Hey, listen, this is valuable, and again, these podcasts, give us an opportunity to really explore topics like IP, which Jason is an expert at. The other side of this is data. So give us a preview to data, because that’s going to be part two of what we talk about.
Jason Novak:
Chronologically, in a development of a digital health and precision medicine company that uses software, uses AI, or even static algorithms, the first thing that happens is data is acquired. That data is acquired to train AIs, develop insights, those insights become what’s patentable, right? So chronologically, data rights is the earliest thing that happens. So it is as important, if not even more important than IP strategy in certain cases. And what we’re going to talk about next time is some of the skeletons in the closets that you may have in your own company, or you may have in a company you’re looking to collaborate with or invest in or whatnot. And some of the things that just simply are not looked at in data rooms, in diligence and due diligence, that could be catastrophic for a company if you’re not really paying attention to where data rights are obtained, if the right data rights are obtained, if the people are giving data rights, even had the right to give you the data. A lot of, this is a very much more of a protective sort of discussion and a defensive discussion, more so than offensive. So it’s, we’re really looking to present watch-outs that you can bring back to your teams, look at your own auditing of your own software, auditing of your own agreements to see whether or not some of these things could be problematic.
Saul Marquez:
Outstanding. Folks, hope you got a lot of value out of today’s discussion with Jason. He is just incredible. Again, partner at Norton Rose Fulbright. In the show notes, we’re going to leave a link to his LinkedIn as well as the firm. If you have questions on the things that you’re working with, maybe you’re missing out. Now’s the time to check it out and learn. Jason, I can’t thank you enough, and I’m looking forward to our next one.
Jason Novak:
Yeah, absolutely, me too. Nice seeing you, Saul.
Sonix has many features that you’d love including automatic transcription software, share transcripts, transcribe multiple languages, collaboration tools, and easily transcribe your Zoom meetings. Try Sonix for free today.
