▶ 0:07:16The subcommittee on environment will come to order without ex without Without objection, the chair is authorized to declare recesses of the subcommittee at any time. Welcome to today's hearing entitled Innovations and Agrochemicals: AI's Hidden Formula for Driving Efficiency. And I recognize myself for five minutes for an opening statement. And first, as I've mentioned to our witnesses, I apologize for the delay in getting started uh this morning. We had a little change in in agenda that we weren't anticipating. It took us a little longer. So, my colleagues will be filtering in here.
▶ 0:07:43Uh hopefully they won't miss much of the opening testimony, but um we will go ahead and get rolling with it. So with that, good morning. Thank you to our witnesses for being with us today. This morning's hearing topic on agricultural innovations is very important to me as my home district is home to over 200 specialty crops, including most of Florida's citrus operations and uh many other different diversified agriculture interests.
▶ 0:08:07I represent what uh we think is probably the largest agricultural district east of the Mississippi, which is a surprise to a lot of people who aren't from Florida and think everyone either lives at uh Disney World or at the beach. But uh we actually have a lot of agriculture uh interest in Florida. Uh we um we we absolutely require safe and and effective access to aggrochemicals like pesticides, herbicides, insecticides, and fungicides which are essential to keeping our crops healthy and productive.
▶ 0:08:35Today we're discussing the current and emerging AIdriven scientific and technological advancements in agrochemicals. We will explore how artificial intelligence is transforming the industry by enhancing key functions such as re research and development, testing, production, compliance, safety, reviews, and applications.
▶ 0:08:53In fiscal year 24, uh I was able to secure $4.5 million in federal funding for the Center for Applied Artificial Intelligence at the University of Florida's uh Institute for Food and Agricultural Sciences GF Coast Research and Education Center. It's a mouthful, but it's our AI center for agriculture associated with University of Florida. That center will serve as a hub for statewide agricultural and AI initiatives and demonstrations with a strong focus on pest management.
▶ 0:09:21And in my district, citrus greening uh has devastated growers and weakened the backbone of Florida's agriculture economy. The research conducted by the institute and other academic partners like the ones represented here today are critical to our discussion. These efforts are not only advancing pest management, but also lead to the breakthroughs that we need in agricultural technology to finally cure the disease that's killing Florida citrus.
▶ 0:09:45Additionally, EPA administrator Lee Zeldon recently announced muchneeded changes to address the backlog of over 504 new chemical reviews and 12,000 pesticide reviews that are well past the expected timelines under the federal insecticide fungicide and rodenticide act FIFA and statutory timelines uh under the Toxic Control Substance Act TSCA as a part of administrator Zeldon's powering the great American comeback initiative advanced ing American
▶ 0:10:16leadership in artificial intelligence as a central pillar with a focus on supporting AI development through clean energy to position the United States as a global leader in AI. I believe this hearing will demonstrate what's possible with AI in agricultural review space and inform policymaking as the EPA continues to develop its AI plan.
▶ 0:10:36I'm eager to hear each witness's testimony and look forward to working with committee members to ensure the United States remains a global leader in AIdriven scientific and technological advancements. I now recognize the ranking member of the subcommittee for his opening statement. Thank you uh Mr. chairman uh for convening today's hearing on innovations in agrochemicals and thank you to our witnesses uh for sharing your insights.
▶ 0:11:06Now, Rhode Island may not be the first state that comes to mind when people think of agriculture, but we're home to uh innovative aquaculture and many small farms and producers, including Wright's Dairy Farm in North Smithfield and Phantom Farms in Cumberland, Rhode Island. Rhode Island's first congressional district is also home to research institutions, scientists, and innovators working on the front lines of resiliency and sustainability.
▶ 0:11:32And in a state uh with a long coastline, finding solutions to the climate crisis is crucial to preventing long-term damage from sea level rise. Agric chemical innovation to introd uh excuse me to reduce the harmful climate impacts of farming through artificial intelligence, data modeling and chemical safety is important and matters to us. AI systems depend on a foundation of long-term highquality data.
▶ 0:12:02For the agricchemical sector to operate safely and effectively, models must account for shifting climate and weather patterns. But that foundation is under attack as we speak. The Trump Musk administration has crippled the National Oceanic and Atmospheric Administration's data infrastructure and hollowed out its scientific workforce.
▶ 0:12:24Dozens of important data sets, reports, and services have been thrown out the window over the past several weeks due to what I believe to be reckless actions. actions that will be uh problematic and cause irreparable harm. Without accurate and transparent forecasting and climate modeling, farmers cannot react and plan ahead. So, let's be clear, no algorithm is better than the data that it runs on.
▶ 0:12:52And if we let politics dismantle the very systems that provide the data farmers use to determine when to plant, water, apply pesticides, and harvest, we're setting ourselves up for failure. Uh and that's across sectors, that's across states, uh that's across our entire country. We must also remember that innovation often begins with research uh at public universities funded by federal dollars.
▶ 0:13:19But these investments are being systematically eroded uh by the Trump administration. The president has caused billions of dollars in federal grants to research institutions and universities. That's going to cause delays to critical work, destabilizing programs, and jeopardizing the very pipeline of talent and discovery that fuels our innovation economy in this great nation.
▶ 0:13:44Our three witnesses today represent the direct benefits of federal uh investments in fundamental research. Thanks to federal funding, scientists of public universities throughout the country are working right now to develop AI models and data science that will make agriculture more efficient. The massive commercial and industrial opportunities available in agriculture today are only possible because of these fiscally responsible investments.
▶ 0:14:13But the Trump administration is not that interested in this research. They're trying to slash the National Science Foundation's budget by $4.9 billion. That's a 55% cut to our n country's steward of basic research. It is the very agency that supports the underpinnings of agricultural technology, including AI. The entities that will most uh that will be most hurt by uh these Trump administration actions are not academic elites.
▶ 0:14:44It's the rural institutions, agricultural extensions in our community colleges, our young students who simply want to pursue and advance the field of science. Thanks to Trump, their opportunities are rapidly disappearing rapidly. We all suffer without federal investment in science to address challenges in a agriculture and climate. Businesses cannot maintain their advantage.
▶ 0:15:11Workers suffer and our global leadership is diminished. Innovation may be the topic today, but the foundation is science. And right now, that foundation is crumbling uh beneath our feet. I urge my colleagues to stand up for our federal scientific agencies and support current and future scientists. We need innovators to continue their work responsibly, uh, ethically, equitably, and well funded.
▶ 0:15:42Uh, so I say thank you to our witnesses. We look forward to hearing from you. And, Mr. Chairman, I yield back. Thank you, Ranking Member Amo. I know Chairman Babin has interest in making a statement. when he arrives, we'll recognize him. But at this point, I'd like to turn it over to the ranking member on the minority side, Miss Lafrren, for her statement. Uh, thank you, Chairman Franklin, and ranking member Amo for this hearing.
▶ 0:16:05Uh, my district is home to the salad bowl of the world, uh, which produces more than half of all the leafy greens consumed across the United States. The Selenus Valley is considered one of the most important agricultural hubs in the nation and its success is due not only to its fertile land but also to the hard work and resilience of the farmers and farm workers who keep uh who keep it thriving.
▶ 0:16:32Agric uh chemicals play a role in sustaining agricultural productivity and food security, but we can't ignore uh possible public health risks that come with widespread chemical use. Chronic exposure to certain pesticides has been linked to cancer, reproductive disorders, and developmental harm. Agricultural workers and communities who live near farms face particularly higher risk because of their proximity to these chemicals.
▶ 0:17:01Today's hearing highlights how artificial intelligence may offer new tools to reduce those risks. AI could potentially help farmers apply chemicals more precisely, target pests more effectively, and ideally reduce the overall amount of chemicals used. Uh that is something that both farm workers and farmers would welcome.
▶ 0:17:25I look forward to hearing from our distinguished witnesses about how these innovations can create safer working conditions, improve health outcomes, and support uh support sustainability in agriculture. However, innovation doesn't happen in a vacuum. AI algorithms depend, as the ranking member just mentioned, on robust, reliable data that is verified through rigorous and systematic reviews of human health hazards.
▶ 0:17:52The fiscal year 2026 so-called skinny budget proposes a substantial cut to EPA's Office of Research and Development Funding, essentially eliminating the very office that researches, evaluates, and provides publicly accessible exposure data for over 1 million chemicals. Housed in OARDD is the integrated risk information system known as Iris which identifies and characterizes the health hazards of chemicals.
▶ 0:18:22Iris is an important source of toxinity uh toxicity information used by the EPA as well as by state and local health agencies as well as international health organizations. The placement of iris in OAD was intentional. It ensures that science is insulated from political pressure. The Iris program is also critical to government efficiency, which should have been something the department of so-called Department of Government Efficiency would have supported.
▶ 0:18:52But unfortunately, this administration seems to be more efficient at gutting programs than keeping our constituents healthy. Even if AI innovations continue adv to advance, we'll lose the ability to make accurate health risk assessments and informed regulatory decisions without a strong scientific foundation.
▶ 0:19:13If we handic uh cuff the scientific arm of EPA, we will discover the consequences of pestified exposure the hard way through impact on people who've been exposed. It doesn't matter how powerful the AI models are if they're based on unreliable or missing data. Using AI to approve new chemicals without the critically necessary scientific health data to ensure they're safe is not innovation. I think that's recklessness.
▶ 0:19:43And with that, Mr. Chairman, I yield back. Thank you, Ranking Member Lorren. Okay. Now, let me introduce our witnesses. Our first witness today is Dr. Brian Lutz, vice president of agricultural solutions at Cortiva Agracience. Dr. Loots is responsible for driving Cortiva's AI strategy to accelerate agricultural innovation. His leadership focuses on leveraging AI to transform the agricultural sector.
▶ 0:20:08He oversees a large globally distributed organization of data scientists and software engineers who are responsible for the integration of AI across various aspects of Cortiva's operations. Our next witness is Dr. Daniel Swale. He's an associate professor in the emerging pathogens institute and department of entomology and nematology at the University of Florida in my home state. Go Gators.
▶ 0:20:29His current research intersects with AI with the goal of discovering and developing novel mechanism pesticides to provide knowledge on chemical nodes modes of action and defining mechanisms of resistances published over 50 peer-reviewed publications. And our third witness is Dr. Boris Camalti an assistant professor in the department of crop sciences at the University of Illinois Orbana Champagne and he is affiliated member of the center for digital a digital agriculture.
▶ 0:20:57His research and extension program focuses on the development of integrated disease management strategies emphasizing the use of artificial intelligence and other technologies to improve precision and sustainability. And I now recognize Dr. Loots for five minutes to present his Chairman Franklin, Ranking Member Amo, and members of the subcommittee, thank you for inviting me here today. So, I'm here representing Corteva, a worldleading pure play seed and crop protection company that's headquartered in Indiana.
▶ 0:21:26I oversee all aspects of digital enablement and artificial intelligence for R&D at Cortetova. And I'm also proud to be from a fourth generation corn and soybean farm in Ohio, northeast Ohio, a town called Lordstown. Let me start with a fact that too few people recognize and that's that innovation lies at the heart of modern agriculture. American farmers are among the most pro productive and produce the most abundant, affordable and secure food source for America and the rest of the world.
▶ 0:21:55American farmers produce 300% more per acre today than they did 70 years ago, allowing greater production from a smaller footprint. This is a great American success story that has innovation at its heart. Today, I'd like to talk about how artificial intelligence can accelerate innovation and in doing so cement American leadership for generations to come. We believe innovation will become more important in the future than it is today because the world is adding more people, but we are not adding more farmland.
▶ 0:22:23At the same time, crops are getting harder to grow as extreme and changing weather patterns place farmers against new and rising insect disease and weed pressures. That's why as a company we invest nearly $4 million every day to in research and development to unlock the latest gamechanging tools for farmers. AI is already playing an important role in three distinct ways. First in discovery, second in development and manufacturing, and third in helping farmers know when and how to use our products to maximize productivity.
▶ 0:22:54Discovery is the process by which scientists identify new technologies that can be turned into products used on the farm. For us, this might mean finding a new molecule that can effectively and safely control pests that harm crops. This is an extremely complex process. We might think of crop protection molecules as keys, and those keys are intended to fit very specific targets in pest, usually proteins, locking the protein and eliminating the pest.
▶ 0:23:20The first challenge is that we don't always know all the locks or the proteins or how they work. So, we don't know what to target. And there can be tens of thousands of proteins in a given pest. This explains why historically discovery has largely been a game of chance. So let's start with the locks or those proteins. For nearly a quarter of a century, computer scientists and mathematicians have been trying to develop models to predict the structure of proteins, but with limited success.
▶ 0:23:45In just the past few years, a new class of AI models has been developed that works with sequences of DNA data to be able to predict the structure of proteins in just a matter of seconds and with accuracy comparable to laboratory methods. Mapping the structure of proteins helps us understand which key we need and how to make it work. So, let's turn to those keys, the crop protection molecules that we develop for farmers.
▶ 0:24:10Not that long ago, we had to screen many thousands of potential molecules by applying them to pests in hopes of finding new ones that might be effective. Today, we're using AI to search the vast chemical universe for molecules that can interact with specific proteins within pests to keep them from harming crops. Our goal is specificity. To find a match that'll do its job and only its job, leaving the rest of the plant and surrounding environment and biodiversity intact.
▶ 0:24:36Bringing it all together, AI has revolutionized discovery by allowing us to trade randomness and chance for prediction, specificity, and design. We can now model proteins and molecules with unprecedented speed and accuracy. For example, we recently used a new AI model to model how 10,000 different molecules might be used in crop protection, all within the matter of just weeks. And the model was able to identify dozens of new potential crop protection molecules that our chemists likely would not have found otherwise. Cortetova is also an American manufacturer.
▶ 0:25:07Our largest manufacturing site is in Midland, Michigan, and we have long led the industry in manufacturing innovation. AI has helped us continually optimize our manufacturing processes. For example, we use fermentation to produce some of our leading products. Over the past few years, we have used AI to both engineer the microbial strains that drive our fermentation reactions as well as have used AI to optimize our reaction conditions.
▶ 0:25:30These improvements have allowed us to run manufacturing operations that are extremely efficient, which in turn have enabled us to maintain a strong manufacturing footprint here in America. Finally, AI is also helping us to better understand when farmers should treat each of their fields based on data related to the specific environmental conditions, predicted pest pressures, and their unique management practices. For example, we're piloting a fungicide timing model combining field specific information with our internal data.
▶ 0:25:58This helps farmers know exactly when to spray to combat key corn diseases. The impact is significant. We regularly help them increase yields by 4 to 10 bushels per acre. Farmers appreciate this because it not only improves the sustainability of their practices, but also puts more money in their pockets. AI is transforming everything we do at Cortetova. It's without a doubt one of the most profound technologies ever to be invented.
▶ 0:26:21We believe there's tremendous opportunity for our government to support and incentivize advanced innovation, including by leveraging the benefits of AI and to benefit American farmers. The race to seize the opportunity has begun, but America is not the only country on the track. If we want to win, we need to move smarter and faster than our competition. As an American innovator, manufacturer, and partner for American farmers, Cortetova believes that with the support of our government, we will do exactly that. Thanks again for having me here today, and I'd be glad to answer any questions you have. Thank you, Dr.
▶ 0:26:50Loots, and I now recognize Dr. Sil for five minutes testimony. First, thank you, Chairman Franklin, and the rest of the committee for the opportunity to testify in this honored house today. I am Daniel Swale and I am an associate professor of insecttoide toxicology and drug discovery in the emerging pathogens institute and department of entomology at the University of Florida. My lab aims to discover and optimize new chemistry for control of insect, fungal and bacterial pests of row and specialty crops as well as insect and aractin pest of livestock.
▶ 0:27:20These pests directly directly lead to billions of dollars lost in economic revenue and for a variety of reasons our ability to control these pest has become severely limited that justifies the need to develop new technologies. The United States has historically been a global leader and has dominated the landscape in innovation and development of novel agricultural technologies including agrochemicals. However, we no longer are the global leader in agrochemical innovation or agrochemical development.
▶ 0:27:50Japan exceeds the United States in the numbers of first-in-class pesticides by over two-fold and China is near equal to the US with total numbers of insecticides produced. Even more concerning, I personally believe the gap between other countries and the United States is expanding rapidly. One of the largest challenges to ensuring economic viability and sustainability of our agricultural industry is the shrinking number of pesticides available for rapid and effective control of pest populations.
▶ 0:28:20I believe the reduced number of efficacious insecticides and the inability to control these pests will be the primary driver of reduced sustainability of our agricultural industry. A prime example of this is the Florida citrus industry. This industry was valued at 9 to10 billion dollars a mere five years ago and is now experiencing near collapse due to an insect named Asian citrusid that transmits a bacteria to the citrus tree to prevent fruit ripening.
▶ 0:28:46Yet control of this insect pest has become a tremendous challenge due to limited number of effective insecticides available. Myers have developed firstin class and and uh highly effective insecticides that kill this pest. Unfortunately, the likelihood these novel chemicals will reach the market is low due to increasing regulatory restrictions as well as an expensive and prolonged developmental pipeline for pesticides.
▶ 0:29:15To remedy this, we have turned to employing AI technologies to discover chemicals of the natural world because the registration requirements for natural products are significantly lower when compared to synthetic insecticides. While natural chemistry is known to have the potential to discover for discovery and innovation of effective of effective molecules, the historical barrier for natural chemistry has been the absence of technology capable of comprehensively understanding chemical structures and interactions at scale.
▶ 0:29:45But the ability to understand this molecular language of the natural world is now emerging. Leveraging advanced artificial intelligence, companies like Invea out of Boulder, Colorado are at the forefront of this transformation. They have developed the capability to predict chemical structures in complex natural mixtures with unprecedented speed and accuracy. Inve f focused primarily on on natural pharmaceutical development.
▶ 0:30:10Uh but together we are applying this innovative AI technology to discover and develop novel agrochemicals. Although high optimism surrounds AI technologies, challenges do exist and these gaps must be addressed prior to maximizing the potential of this approach. The outputs of AIdriven discovery are only as good as the inputs you provide the system and currently the inputs needed for agrochemical discovery remain poorly understood.
▶ 0:30:39To date, agrochemical discovery and earlyphase development has been restricted to the private sector which do not share data sets due to intellectual property restrictions. This has led to gaps in public knowledge needed for appropriate AI inputs and I believe has limited the utility of AI for discovery of new aggrochemicals. So how do we address this challenge? I believe AI enables public entities such as universities to play a key role for the very first time in the discovery of agrochemicals.
▶ 0:31:08This is due to their ability to test fundamental questions oftentimes not addressed by the private sector. Thus, involvement of the public sector into agrochemical discovery will allow fundamental data sets and AI inputs to be public that will promote advancement and innovation at the national and global levels. I encourage Congress to consider investments into academic institutions that can perform agrochemical discovery which will promote public private partnerships at the state and national level.
▶ 0:31:38It is here where we will find innovation and advancement of the agrochemical industry to restore the United States at the top of the podium for agricultural innovation. Thank you for the opportunity to contribute to this conversation and I look forward to our discussion today. Thank you Dr. Swel, Dr. Kamalleti, you're now recognized for five minutes testimony. Chairman Franklin, ranking member ammo and members of the subcommittee. Thank you for the opportunity to speak with you today.
▶ 0:32:02I'm Boris Camilletti, assistant professor in the department of crop sciences at the University of Illinois Urbana Champagne and I am also affiliated with our center for digital agriculture. My research focuses on improving plant disease management by using artificial intelligence, remote sensing and other approaches to help farmers use fungicide more efficiently and sust sustainably. Originally from Argentina, I came to the US as a fullbray scholar to study sustainable crop protection.
▶ 0:32:30Today I lead a multi-disiplinary team combining plant pathology, remote sensing and AI to address one of the most urgent threats facing US soybean farmers, red crown rod. Red crown rod is a soyborn disease that was first detected in Illinois so fields in 2018. It can cause premature plant death and yield losses of up to 50%. It's spreading rapidly across Illinois, which remains the hardest hit state, and has also been confirmed in fields in Indiana, Kentucky, and Missouri.
▶ 0:33:00Because it survives in the soil for several years, it's extremely difficult to manage. Some farmers are even switching to continuous corn to avoid the disease, which bring its own sustainability challenges. What makes red rod especially challenging is that it's hard to detect in its early stages. The disease shows up in scattered patches and symptoms often don't appear until late in the season. The scouting in the field is labor intensive and impractical at that large scale.
▶ 0:33:28That's where artificial intelligence and remote sensing come in. My team uses satellite imagery and machine learning to identify red ground hotspots in the field. We train the models with high resolution spectral data from visible to near infrared bands and use ground truth observations to teach the algorithm what disease plants look like. These models can detect disease areas and just as importantly they can track disease progression over time.
▶ 0:33:56This technology has real on farm impact. We are building tools that generate prescription maps. So instead of applying fungicides across the entire fields, farmers can target only the infected areas. For example, if only 25% of a dis of a field of a field is diseased, our approach could reduce fungicide by up to 75%. This strategy reduces cost, protects the environment, and help preserve the efficacy of our fungicide tools.
▶ 0:34:25We are now preparing on farm trials to compare traditional uniform applications with AI guided sight specific treatments. These trials will help us measure the yield, disease severity and chemical use side by side. So we can show growers how what precision disease management looks like in the real world. Beyond soybeans, this technology has raw potential. It can be adapted for rust diseases in corn and wheat for tree crops like almonds and pistachios and even for weed detection.
▶ 0:34:55is a platform for precision agriculture that reduces chemical use while improving control. This work is possible thanks to the collaborative environment at the University of Illinois and our center for digital agriculture. This center brings together scientists and engineers across disciplines to solve real world problems in agriculture. It provides the infrastructure, computing power and partnership we need to develop these tools and bring them to farmers. I also want to highlight the importance of land grant research at the University of Illinois.
▶ 0:35:25Our mission is to serve the public. The AI technology I've described today were made possible by public investment including Hatch funding and support from farmer led checkoff programs such as the Illinois Soybean Association. PL public support ensures that the tool that we developed serve the public good, helping farmers make better decisions, lower input cost, and manage diseases with greater precision and less environmental impact.
▶ 0:35:51In closing, I want to emphasize that AI is not just a buzz word in agriculture. It's a practical solution to a real problem. With regrown rod, we have a narrow window to act. AIdriven tools give us the ability to detect the disease early, respond quickly and use chemicals more responsibly. This is the future of crop protection and your support makes that future possible. Thank you and I look forward to your questions. Thank you Dr. Cameleti and I thank all of our our witnesses for their testimony. I now recognize myself for five minutes of questions.
▶ 0:36:24Dr. Swell, some for you to begin with and for all of you. There's so much to unpack in your your written testimony. I know five minutes didn't give you long to to get into that. So I'd like to dive a little deeper. Dr. Swell, you mentioned your work in developing a dual mechanism insecticide to fight the Asian citrusid which spreads citrus greening. Obviously a massive issue for us in the state of Florida. From your perspective, how close are we to having a new commercially viable insecticide derived from your research that could be used by citrus growers in Florida? And and what do you need?
▶ 0:36:53What kind of support uh and or partnership do you need from the federal government that would help accelerate that? That's a great question. Um I I think that we have commercially viable compounds that that can contribute to uh the decline of citrus. Um unfortunately the these are synthetic insecticides and the regulatory pipeline is prolonged and unbelievably expensive. Uh so the an direct answer to your question is I I don't think we're close at all even though we have the tangible compounds in hand.
▶ 0:37:22Um and that is why I'm very excited about opportunities uh with companies like Invea because identification of natural compounds can mitigate this prolonged and arduous process that ultimately likely usually leads results in failure. Um so if we're able to identify natural compounds um that we can potentially have uh something that can be field ready in a few years.
▶ 0:37:48So, for those of us that aren't deep into the science as you are, can can you explain a little bit about how AI specifically will help with those natural um um alternatives versus the synthetic? I mean, why is why why is it Uh yeah, there's a there's a there's a number of differences and my colleague here from industry might be able to expand on it a little bit, but uh the natural world is is largely untapped due to the complexity of one the chemical structures and how you obtain these mo how how do you get the molecules out of the
▶ 0:38:18plant. Um AI is contributing to that quite significantly. So AI just what will assist in the speed just because of the the vast speed of being able to understand what this molecular language actually looks like. natural compounds are uh very very bulky and huge and look funny uh compared to synthetics. So how do we parse that down and utilize this? Uh we haven't been able to wrap our head around that but it appears as though artificial intelligence has begun to pick up ground on that. Right. Dr.
▶ 0:38:48Lutz, could you comment on that? Yeah, absolutely. Um so uh to the point made by Dr. So I mean natural products, naturally inspired products have a greater complexity than synthetic molecules. Synthetic molecules tend to be small in size. Biologicals and natural products are much larger um and more complex.
▶ 0:39:07So not only with the speed but um when you think about the the vastness of chemical space um it is uh almost impossible for us to understand as humans how many potential molecules could be out there to be able to interact with pests to prove as effective crop protection molecules. AI models allow us to work through this space at speeds completely unprecedented. Um at Corttova we have been moving into this space to Dr.
▶ 0:39:35SW's point um in part due to some of the regulatory considerations with the amount of time it takes to get new synthetic molecules um approved. So we see moving into this space as an area where we can help bring new and effective innovations to farmers uh faster but it is an extremely complicated space that AI is letting us uh distill. Okay.
▶ 0:39:56You'd mentioned in your testimony that um Cortetova can is using AI to model structures of over 10,000 molecules and um you can do that and how and look at how they might be used for crop protection in a matter of weeks. How does that compare to what you'd be doing traditionally without the use of AI time wise? Yeah. I mean um discovery traditionally is like finding a needle in a hay stack.
▶ 0:40:19I mean, you're talking about screening tens of thousands or hundreds of thousands of molecules before you find one in which you're willing to make a bet on to be able to go through the perhaps decadel long process of trying to get it uh registered. Making that decision early on is extremely important. And so our discovery funnel um we're searching through this space as best we can to find where we want to make those bets.
▶ 0:40:44AI opens up that funnel and lets us find what we believe are more impactful, more efficacious, more sustainable molecules sooner that we want to make those bets on as we move forward. Right. Thank you. I have other questions, but I'd now like to recognize the ranking member from Rhode Island for five minutes of questions. Thank you, Mr. Chairman. Um, in late January, the Trump administration ordered uh the Department of Agriculture to remove all climate related data uh from its websites.
▶ 0:41:13Uh, this politically motivated data purges, denying farmers access to vital information and assistance they need to apply for federal loans, receive up-to-date technical assistance, and make time-sensitive crop decisions. Without these critical data sets, farmers are left in the dark and the consequences for our food system uh could be severe. Following a lawsuit just last week, the Department of Agriculture agreed to restore climate change related content to its resource pages.
▶ 0:41:43This is yet another example of the unlawful and efficient and completely unavoidable harm that we've seen from this administration in in its uh engagements that have uh really hurt many Americans. So, I've got this question to every witness.
▶ 0:41:59Can you speak to how federal data sets support or enable research into agrochemicals and how would research or innovation capacity be impacted if access to publicly uh available data were uh reduced? I'll start with you Dr. Lutz. Yeah, I mean it's very important to understand um sort of the chain of what we do in order to deliver AI enabled technology to farmers.
▶ 0:42:27Not only does it apply, not only does it deal with how we apply models to finding those solutions, but it requires on the models themselves in which we have depended on very important industry partnerships to develop those models. Those models can actually then be traced back to basic research that has been done at our universities for the last half century that has set the stage for how we build these models.
▶ 0:42:50And of course the models themselves sit a top a data foundation that we absolutely could not develop um AI models without. Um we do uh depend on the many data sets that the government has supported over the last several decades um to be able to drive innovations uh for farmers. The fungicide timing models that I've mentioned that we have been developing do rely on climate data um that we get from the public sector to be able to deliver those.
▶ 0:43:19Uh yeah, I um I'll say that I think federal funding has been a foundation for which major breakthroughs and innovation is built. I personally believe that a commitment in fundamental science by the federal government is is critical for the US to remain at the leading edge of innovation and advancement. Uh you know our university has asked the question of of why and hows and without an understanding of why and how a system works it's challenging to innovate. Dr. Camelar. Yeah.
▶ 0:43:46So I think you highlighted very well that in order to train this model we need to have very robust data and to do that in our case for example we need to go and sample as many fields we need to travel we need to go to every field to record data to train these models and that is done with workforce with the students with possum so specifically for this project we have some proposals that were going to the USDA and those programs were put on hold that is
▶ 0:44:16impacting us our workforce in our ability to go and collect real good and robust data to train these models. We rapidly found another program at the NSF with the cyber physical uh systems where we will be submitting this this this proposal in order to support our students and post. Yeah.
▶ 0:44:38and and you know look I I think it's important to hear from you directly about how foundational uh the the partnership with uh federal funding your institutions uh supporting uh the private sector.
▶ 0:44:52there's a a robust amount of partnership that uh unlocks possibility that gets us on the first steps of innovation and and again we we have seen action that cuts right at that uh and undermines our our capacity to uh to grow.
▶ 0:45:10uh you know a recent uh Forbes article estimated that that uh the Trump administration's proposed National Science Foundation and National Institutes of Health cuts uh will cost not save American taxpayers $10 billion annually. These cuts dismantle our pursuit of basic research that has consistently generated strong dividends for the public. Every dollar in basic research generates about $1.40 to $210 in economic output.
▶ 0:45:39And I I believe that's a conservative uh uh estimate. Uh and without this basic research, we will not have the information needed for industry uh to perform applied research and innovate new products and processes. And so I my my time is is set to expire, but I I do want to to echo what you have said both in uh your testimony and response to that first query that we need this partnership. We need that foundational data in order to get this work done. And with that, I yield back. Thank you.
▶ 0:46:09I now recognize the member, the gentleman from Alaska, Mr. Beggish, for five minutes. Thank you, Mr. Chair. Um, first question to Dr. Loots. While AI has greatly accelerated uh proteomic target discovery, how is AI being used to ensure that we don't see unintended consequences of such uh deployments?
▶ 0:46:31Yeah, I mean with any new technology um that yields a lot of opportunity like AI, we also have to look at the other side of the coin which are the risks that it can bring as well. Especially when we look into the biological and the chemical space um models if used inappropriately or trained inappropriately can be used to create technologies that are um of not good intent.
▶ 0:46:52Um we uh strongly encourage that all players in industry or elsewhere that are creating these models take precautions to not train models with data um that potentially makes them available for inappropriate use. uh we also have to put in place simple and common sense I think uh frameworks to be able to uh do testing red teaming uh triing of models to assess their ability to uh produce dangerous outputs.
▶ 0:47:23Um we also need to have clarity I think in uh the regulatory agencies that can oversee the uh the use of these models. Right now it is a bit of a patchwork. Um, I also would caution us against allowing um this to go to too uh too local of a level where we'd end up with a patchwork uh set of regulations that might be established um at the state level that can be very difficult to navigate how these uh these models are used um more broadly.
▶ 0:47:52We certainly have to look at the safety um of these models, but we just need um any regulations to to uh not hamper um the ability to to innovate with them. Ju just to get a sense and this is a this is open to to anyone uh today. Um how much do we think yield per acre might be improved through AIEL agriculture management practices? What's what's available to us? What's at what's at stake? It's considerable.
▶ 0:48:21I I mean I'll start but the um if you if you start thinking just about yield potential and I'll use US corn as an example we know that we can produce on the best yielding acres over 600 bushels per acre and yet national yields are around 180 bushels per acre much of that loss is driven by controllable factors many of which can just be uh mitigated by using products much more effectively so applying at the right time to avoid losses um but then when you go beyond that the ability to leverage AI to
▶ 0:48:52identify attributes that we can then improve crop genetics, the seed genetics. There's tremendous additional potential there. And then of course once we get to the crop protection products themselves um to make them more specific and more efficacious. Um truly we're at a point where I believe whether it's um acute or targeted um impacts that are affecting agricultural production in any of your states um or for our ability to meet the growing demand of agricultural globally.
▶ 0:49:20Um there is very little that these technologies can't solve or or um help support. Yeah. I'm going to defer to my colleague here. That's a little downstream of me. I work upstream in discovery and early phases of development. So yeah. So in soybeans specifically in our case, we are dealing with a disease that is taking up to 50% of the yield in soybeans. That's a deal breaker. So some farmers are not doing soybeans anymore because of this economic impact.
▶ 0:49:51So we have a huge window to improve our management and to improve the yield directly. But the other thing is if we want to be profitable, we need to think also in lowering the the inputs cost. Um so if we can reduce or make a more efficient use of fungicide is less money that we need to invest in a fungicide to keep our yield. So in that way we can increase our revenues or returns. That's pretty remarkable response there. So we could we could double soybean output.
▶ 0:50:20We could triple more than triple uh corn output per acre. And that's not even touching the input genetics. that's just optimizing what we have right now. That's really impressive and I think it highlights just how important it is for us to make these investments in AIdriven uh productivity gains uh in agriculture because as productive as the American farmer is, they have the potential to be much more productive going forward. And with that uh I yield the balance of my time. Thank you, Mr. Begitch.
▶ 0:50:50I now recognize the ranking member of the full committee, Miss Lorren, for her five minutes of questions. Well, uh, thank you very much, um, Mr. Chairman. This has been an interesting, uh, hearing and I think, you know, there's a lot of exciting opportunities here, uh, with AI. Um, my, um, in my district, uh, there are a lot of innovation going on in terms of, um, pest management.
▶ 0:51:16For one thing, one of the things that I saw recently was uh and this works for weeds, not necessarily for pests, but it's uh laser zapping of weeds. And it's a big machine and it's got more computers than a Tesla and it identifies what's the crop and what's not and and with lasers eradicates the weed. And it's better than chemicals because it incinerated that weed and it's not coming back.
▶ 0:51:45So there we have a solution for some of this but some of it we don't in terms of uh pesticides. I recently visited a wonderful group of nonprofit called Jacob's Heart in Watsonville, California. And what they do is provide support to families where there's a child with cancer.
▶ 0:52:07And up on the wall they had a little map of the cancer uh clusters and the childhood cancer clusters were very much in the a areas and not in the urban areas and it it does um reinforce the concern that exists about health exposures to some of these uh chemicals.
▶ 0:52:29So, the idea that we could find uh new uh efforts to to fight these pests that were less harmful is really an exciting uh opportunity. Here's my question. You're using AI to find uh to unwind proteins, to find new approaches, but you could also use AI to evaluate the health risks.
▶ 0:52:55However, I'm wondering about the database that's available to you that's proposed to eliminate Iris. I don't know whether that's sufficient now to utilize AI uh to test what you found against health risk or what could be made available. Anybody's got an answer to that question, I' I'd love to hear it.
▶ 0:53:19Um well, so first I I would just um respectfully disagree with the idea that there are hot spots of uh of cancer near the use of pesticides. Um I'm not familiar with the the study or the data that Well, it's not a study. I just looked at the map of their clients and it just jumped out at you. Yes.
▶ 0:53:40But I do believe any of the products that we produce, you know, we go through rigorous safety testing and when used as directed um you know we find no evidence of any uh systematic uh effects with with human health. Um to the broader question though around how can AI um help with the digestion of additional data sets including data sets related specifically to human health or any other attribute related to the safety of the products that we we produce.
▶ 0:54:08there's a lot of opportunity to apply um AI in these ways. Um to your point though, it it very much depends on the data and the data that are available and having reliable unbiased data to do that and and how do we get additional data? And the other question is AI systems hallucinate sometimes. How do we guard against that in an area that is so important as this? I don't know.
▶ 0:54:33Uh doctor, do you have a answer to that or Uh I I personally think that over reliance on technology um can put us can kind of shoehorn us into to a problem, right? If we're not if it's premature, uh then then we have issues. And I think the human pharmaceutical industry is a perfect example of this. Um back in the 2000, they they sequenced the human genome and we were going to have the answers to all of the problems, right?
▶ 0:54:59We had all of the perfect drugs, uh and we were going to be able to make things cheap and have all chemicals for all diseases. But that actually was was opposite of the case. Everything became more expensive and more prolonged with the new technology. If I may, Google recently unwound unfolded every protein in the human body and posted it for every researcher in the world to look at. It's great to have the human genome project. We supported that out of this committee.
▶ 0:55:28But without that advance, there was um a limitation on what could be done with it. And I think there's a lot of reasons why pharmaceuticals are costs through the roof, including some that are I don't agree with President Trump on many things, but uh they are being uh the costs are being reduced in other countries and we're covering the cost of of the research.
▶ 0:55:54So I don't think that's a direct well I would enjoy the opportunity to continue with you because I would because there are many of studies about showing the drivers of of what happens when you overrely on technology. Ultimately it comes back um but you have to have the databases and thus the inputs to inform that model. I see my time is expired Mr. Chairman so I will yield back. Thank you Miss Lafer and I now recognize my colleague from North Carolina Mr. Rouser for five Well thank you very much. This is a very interesting uh uh subject.
▶ 0:56:25Uh so in North Carolina, agriculture is a greater than hundred billion dollar uh industry and u uh our largest industry by far. Uh but you also have the effect of uh so many folks moving into the state. um fields that used to grow tobacco, sweet potatoes, soybeans, cotton, uh you name it, are now growing houses and and those houses are are there to stay.
▶ 0:56:53So, you got to uh in order to, you know, feed our country and and and feed the world, we've got to produce that much more on less. Uh can you talk a little bit about uh AI and and the role uh in achieving that end? Yeah, and maybe I would start. Um I was one of the transplants actually from Ohio to live in uh the great state of North Carolina for several years um and and loved my time there.
▶ 0:57:18But I do understand um that there is a competing demand um for things like housing and other development with agriculture. Um I will say that several years ago the narrative was how are we going to feed the world? Um and and it's a true concern. My personal concern is quickly eroding of whether or not we can do that because of the types of technological advances we're seeing. I absolutely believe that this is possible.
▶ 0:57:45Um I believe the challenge is is just making sure that we can get those technologies to market. Um because the points that we were making earlier, the ability to increase productivity per acre is still quite tremendous. Um you know while I think we need to have some thoughtful balance and different competing needs for land um I don't doubt our ability to feed the world um in coming decades. So the uh efficiency of application um is that much greater to achieve that or give me an example there.
▶ 0:58:14It's it's not only the efficiency of input use, right? So, making sure that we're using inputs in order to not have a limitation to production to to maximize production, but we also are developing the next generation of technologies that have seed that can withstand more stresses as well as crop protection products um that are are much more efficacious. But altogether, we we certainly have the ability to continue to drive productivity per acre much higher than where we are today.
▶ 0:58:40Now you you hit on something that's uh uh I think more true today than it has been in the past or at least it seems so is where say you start off the year um it's really really dry and hot uh you hard hard to get the crop out of the ground.
▶ 0:58:59uh or the con converse where you know it's really great the first part of the year you get the crop out of the ground and then you just have baking sun for uh two three uh you know months very very dry. Um what what's the potential for for lack of better terminology of weatherizing these crops where you have a more uniform crop based on very erratic uh weather patterns? Yeah, that's a great question.
▶ 0:59:23So when we are developing new products, especially on the seed product side, we not only look at the genetic potential of the seed, so it's top-end performance, but we also closely watch things like what we call yield and yield stability. Um, yield stability is very much looking at the ability for a crop to withstand those sort of acute stresses.
▶ 0:59:42Um, and when you've looked at the last several years, I mean, actually, if you look at USDA production data here in the US for corn and soybeans, it's been quite remarkable where, of course, we have local events that are very hard for farmers to get through, but in aggregate, we've actually been able to get through many of these um sort of severe uh seasonal impacts uh with with with less impact to productivity than what we have been able to historically. And a lot of that's due to the resilience that we're we're um integrating into crops.
▶ 1:00:12Yeah. Uh so there's some concern out there about the use of pesticides and herbicides and how that affects the food supply. Any comment from any one of the three of you with in regard to that? I I mean I would just emphasize again that look, we don't want to produce any products that aren't safe for humans and safe for the environment. and we stand behind all of the science um and the riskbased analyses that we do um of our products including with our government agencies and regulatory bodies.
▶ 1:00:40Um you know I I think that we can't lose sight of the fact that um without uh our crop protection products today production would be much lower and we already have over 800 million people around the world that don't have enough food to eat and we're adding two billion more people um by midentury. Um it's just not possible to imagine a future where we can feed everyone without these tools for farmers. Yeah, absolutely. And very well said. I yield back my time. Thank you, Mr. Chairman. Thank you, Mr. Rouser. And I'll recognize my colleague from Oregon, Miss Bonamichi. Thank you very much, Mr.
▶ 1:01:10Chairman and ranking member, but really thank you to the witnesses for being here and for your expertise. So, I represent a district in the Pacific Northwest. Agriculture is a significant part of our economy. We have a a lot of plant nurseries, wineries, uh hazelnut farms, uh in Tamote County, which is in the district I represent. Of course, we have dairy. Uh statewide, we have a lot of wheat, beef, but I I want to say that the research-driven agrochemicals are really an important part of what keeps industry productive and sustainable.
▶ 1:01:41uh scientists at some of our Oregon universities have been studying sort of emerging issues like agrovot voltaics where they're uh locating solar panels and agriculture on the same land showing really promise in reducing chemical inputs but also improving water efficiency and lowering energy costs. So innovations like that we should be scaling not sidelining. I want to note that uh concerned of course about the recent staff cuts at the agricultural uh research service.
▶ 1:02:09um just disrupting some of the work in Oregon, but also on critical issues like weed control, specialty crop breeding. Uh those are concerning for me. Um I look forward to exploring in this committee how we can strengthen this ongoing innovation and protect public health and the environment and support farmers.
▶ 1:02:26But I I do want to note particularly as a member of the bipartisan AI task force that I certainly recognize the potential tremendous potential with artificial intelligence, but we have not yet address the privacy, security, and bias issues. Uh we wrote a report. A report's not policy. So we don't have that federal policy yet. And I know you're concerned about 52 different or excuse me, 50 different standards, but we have to have something in place at the federal level. Um Dr. Swill, you mentioned that the outputs are only as good as the inputs. I agree with that.
▶ 1:02:56We have workforce issues as well. And I also share the concerns raised by several of my colleagues about cuts and layoffs at science agencies, especially Noah, uh the National Weather Service, um some of the pesticide reduction work I did at the state level. I know that weather forecasts are really critical for application of pesticides. And Dr. Swell, I noticed you had grants from NIH, NSF, USDA. Um, also the uh what's happening with our I'm on the education committee as well and the higher education subcommittee.
▶ 1:03:24I'm extremely concerned about attacks on higher education uh especially land grant universities. Thank you for mentioning that. Also, thank you Dr. Camei for being here as a wonderful example of the contributions immigrants bring to our universities and to our country. So, Dr. Dr. Cameli, how would the federal workforce reductions at USDA like those that are affecting weed control and especially crop research, how would those affect our capacity for additional development in AI supported agrochemical work?
▶ 1:03:55Sorry, I I don't know if I understood correctly the question. Well, it we're seeing workforce reductions at USDA. Yes. Especially um I'm concerned about um reductions and the the work that USDA does on weed control, specialty crop research. How would these reductions re uh take away from our ability to do more innovation in this area? Well, specifically in our case, we collaborate very closely with the USDA and was affected by all this situation.
▶ 1:04:24Uh but again, it's affecting the the workforce. We don't have any well now we don't have any students or posto that can help us to collect the data that we need from the fields to be able to improve these models. These models rely very heavily on the type of data that we get and they need to get more and more data every season in order to be improved and that's needs more workforce.
▶ 1:04:46So h in that particular situation is less collaboration projects that we have together uh to support this and we're hearing the people who are graduating with degrees in science fields and engineering are being recruited by other countries. I mean it's with these the cuts at our science agencies it's really concerning. Dr. swale. I think about my my uh producers, small plant nurseries, uh the smaller specialty crops.
▶ 1:05:11They might want to adopt some precision tools to help them reduce um agrochemicals, but there's costs and and training barriers. Um how can academic research help scale AI technologies can benefit these small as well as the larger producers and what successful extension or public private models would work with the specialty crop sectors? Uh this is a great question.
▶ 1:05:33I I you know I think that AI allows and I see we only have 30 seconds so I'll be brief but I think you know AI allows universities the opportunity to participate in the discovery pipeline all the way from the beginning to the end and it it it we have all of the pieces in place. A lot of universities University of Florida included have uh supercomputers and major investments in artificial intelligence um as well as the expertise to be able to to to design these molecules.
▶ 1:06:01Um what we really need is um to be able to leverage the collaborations that we built with small startup companies such as Invea uh that can kind of put all of these pieces together and actually begin to hit uh to to exploit the technologies that are there. Thank you. And as I I know I'm out of time, but as I yield back, I just want to emphasize again the importance of the basic research done at our universities across this country that leverage dollars in the private sector. And I yield back. Thank you. All right. Thank you. I now recognize our representative from California, Mr. Fong.
▶ 1:06:32Uh, thank you, Mr. Chairman. Thank you for allowing me to join you all on this uh this vitally important topic. I don't serve on the subcommittee uh but uh the chairman was gracious enough to allow me to wave on. Um I am I represent the food production capital of California, the Central Valley. Um and so um you know, this is a very important topic. For multiple growing seasons, cotton farmers in my district have been hamstrung by bureaucratic red tape preventing them from using otherwise effective aggrochemicals to combat uh the extremely destructive lag pest. Uh like the chairman, I have uh I have citrus.
▶ 1:07:02This the Asian citrus sillet is a grave concern to us. I've got grapes, almonds, we have the glassy wing sharpshooter. I go to the laundry list of of of pests that we're trying to deal with with the certification process for EPA labels taking upwards uh of 13 years to complete. Uh it's imperative that we we streamline the process uh from R&D to application to ensure growers have have access to vital and safe pesticides that allow them to produce affordable food for our nation. As infestations have only grown year-over-year and Mr.
▶ 1:07:32Luch, you're veryware aware of this. Um you know, growers are are without the effective um tools are now using larger amounts of less effective products which is not only costly but doesn't solve the problem. We're killing beneficials. uh we're leaving growers with an ongoing infestation. Um I mean this is is it's just it's a horrific uh situation when we um so that's why uh in your testimony, Mr.
▶ 1:07:57Loots, you you you shared how AI is accelerating the new product discovery and development process u bringing even more effective pest pest management tools to the market. How can Congress restructure the label approval process to ensure that these effective new new tools can get into the hands of growers faster? Yeah. Um thank you representative Fong. I mean it's a great question. Uh it's probably one of the most daunting things that weighs on my mind.
▶ 1:08:20Um you use the example of some of the challenges faced in your state especially around um cotton with the lias issue. Um this is what makes it particularly challenging right because we actually there's a a great technology in the market already um that can help uh mitigate a lot of those losses.
▶ 1:08:36So as we think about our opportunity to develop uh new technologies um it becomes uh a bit challenging right to think about um the the requirements to make those investments to bring new technologies to the market when um when when existing technologies um are already blocked. I think what I would emphasize most though as we think about labeling and registrations and again you know we as a company do not want to produce anything that is not um safe for people, safe for the environments um safe for farmers.
▶ 1:09:05Um but we also recognize right now the registration process is um terribly long. Um anything that we can do um to streamline uh the registration process uh would would be extremely impactful for our ability to deliver new innovations. The only thing I would emphasize is that we wouldn't necessarily trade speed for uncertainty. We need to have a regulatory process that we know how to navigate where we um know how to get things to the market. But I believe we can do it much faster. Yeah.
▶ 1:09:35And and this is the important piece as AI allows us to screen and develop these new um tools. Uh we have to get into the market and that's the that's the the tension that's that's there. Um we've seen it in California. I'll tell my colleagues, um, the state of California, uh, allowed farmers instead of using a tool that they knew worked, was using a tool they didn't think worked, they they thought would work, but it didn't, and it it killed beneficials and it led to more pest outbreaks, and that's something that we want to avoid.
▶ 1:10:02Um, uh, I want to ask Dr. I want to ask Dr. Camei, uh, pesticide and biological applications are costly for growers, uh, which makes the need for precision application all the more important. uh your research uh in terms of AI and satellite imagery for targeted intervention. Um how can both growers and pesticide manufacturers leverage AI to improve the efficiency of their operations through precision application timing and methodology? Yeah.
▶ 1:10:30Um I used to live in near the Fresno area so I'm very familiar with the central valley. H in that case we had some diseases in pisio orchards for example and I think AI has a strong potential to identify h this disease before that we can see it with our eyes that is because there are some changes in the canopy that we can detect with remote sensing that we don't see with our eyes that is the key because that can will affect the timing of the fungicide applications not
▶ 1:11:00only to control the disease we have it but in order to not to spray fungicides If we don't have the conditions for the disease, we need to reduce exposure to pesticides in uh for people working in in those orchards. And I think the AI is a great tool to anticipate to predict disease outbreaks and more importantly to avoid unnecessary use of fungicides in that case. That is my field of study. I appreciate that.
▶ 1:11:26Uh allowing more precision uh applications, spraying less, being more effective. This is an exciting time and we can use this technology to u to make u make things better for our growers. Uh with that I yield back. Thank you Mr. Fong and now I recognize my colleague from North Carolina Representative Ross for five minutes. Well thank you Mr. Chairman and thank you to all of the panelists for being with us today.
▶ 1:11:51Um, as you know, innovation in the agrochemical industry has transformative potential not only for the United States and global agriculture, but also for institutions leading the charge in research, development, and education.
▶ 1:12:08North Carolina State University in my district, a recognized leader in agricultural research, is harnessing the power of new technologies, including artificial intelligence to advance sustainable and efficient agricultural practices through AI powered data analytics. The university accelerates the discovery of new agrochemicals, optimizing formulations to improve pest control, disease management, and crop protection.
▶ 1:12:37Over the past century, North Carolina has grown to become a national leader in agriculture. It is a pillar of our state's economy. Some agricult agrochemicals used on our farms and across the United States have had a significant environmental and public health impact, not always to the good.
▶ 1:13:00Long-term exposure to harmful chemicals, particularly from non-organic food products, has raised concerns regardly regarding the safety of both consumers and those involved in food production. My hope is that by focusing on the development of environmentally sensitive agrochemicals with the assistance of AI and emerging technologies, we can protect our environment, public health, and farm workers from adverse effects while maximizing food production.
▶ 1:13:31Sustainable agriculture depends on maintaining the health of ecosystems, the safety of food, and the well-being of farmers and workers. A recent survey by the Association of Science and Technology Centers has revealed that US adults value science but aren't necessarily connecting federal investments in science um that to things that are useful in their daily lives.
▶ 1:13:58Each of you has testified how federal investment at university research levels has enabled the opportunities with AI and agriculture today. However, the Trump administration has proposed to cut much of this funding, including a 55% cut at the National Science Foundation. NC State is the largest recipient of National Science Foundation grants in our state.
▶ 1:14:24These cuts will our universities in their ability to perform this cutting edge research and train the next generation of researchers. Dr. Swale, in your testimony, you note that the US is no longer a global leader in agricemical innovation and development. Can you elaborate on the factors that have contributed to this decline and what steps we can take to restore our leadership in this space? Yeah, that's a great question.
▶ 1:14:52Um, and I I think there's a number of different reasons for this. I think if I was to put pick one for for the sake of time, it's just the simple cost that it takes us American-owned companies to move a molecule from my lab into his company and out to the farmer's hands. And then the 10 to 13 years in the estimate that it is for that. I think what that has driven is more meto science. Uh this company has already developed this chemical, so we're going to do something very similar. Um and there's no innovation.
▶ 1:15:22um or we're just going to repurpose existing chemicals and change it slightly. Um I think that that has reduced the innovation. So I'm hopeful that the inclusion of artificial intelligence platforms allows us to, as he's said, you know, we've mentioned a lot of times map the protein structure and find these molecules, but that's only one part of it, right? We don't know what these chemicals look like. Um what does the molecule look like to kill a fungus or a sucking pest versus a chewing pest? And all of the chemical parameters are different.
▶ 1:15:51Uh so can we in academic science can we start to define these inputs that right now are isolated within these large agrochemical companies and why that's relevant is it allows us the opportunity to start focusing on more niche uh agrochemical agricultural pests right so um the the large um and they do a fantastic job this isn't to mitigate their efforts but they focus on global e large economically relevant pest and citrus doesn't make the cut um a lot of a lot of these other more
▶ 1:16:22commodity uh products don't make the cut. So I think moving this utilizing AI um allowing us to start to develop products towards more of a niche niche environment to expand innovation in the United States. And of course to my point uh my intro point NSF funding frequently helps with that because NSF takes the risk on some of the innovative uh research and some of these niche areas. Thank you Mr. chairman and I yield back. Thank you.
▶ 1:16:52And uh I thank the witnesses for their valuable testimony today. Members for their questions. Oh, I'm sorry, Dr. Babin. I didn't see you slide in there. To bat cleanup, our chairman of the full committee, uh Mr. Babin from Texas. Sorry, I'm running a little late. Mr. Chairman, I want to say thank you to the witnesses for being here as well. Um Dr. utes Cartava is a leader in agrochemical innovation and as we heard from your testimony, America should seize the opportunity to lead.
▶ 1:17:21The Trump administration has made it a priority to invest billions of dollars into this new age of AI. We want to ensure that the private sector is not hampered by bureaucracy and red tape, but we also must ensure that we're getting the science right across the board. And uh number one, how can Congress and the federal government best support the private sector to bo bolster this Yes. Uh thank you, Mr. Chairman.
▶ 1:17:51Um you know, one of the things that we we can't lose sight of, especially when we think about AI driving innovation in this space is that it is a um complicated ecosystem. We as Corteva are applying these models to drive new innovation for farmers. But we have to recognize that many of these foundational AI models have actually been built by the tech giants, American tech giants that are training these models um at at tremendous scale.
▶ 1:18:16They they uh not only benefit agriculture, they benefit other life sciences, human health, pharma as well. The advancements in those models build off of decades of research by brilliant minds at our university that have been pushing the forefront of of how AI can um really be be used. I mean the the big tech companies now have just made the factories effectively in which these models are built, but they're built off the thought leadership that come from our universities.
▶ 1:18:43And of course, the fuel for all of these models is the public data um that and the private data that goes in to feed um and train them. I think what's really important to recognize is that if we take any one of the legs of these stools away, America can lo can we can lose our competitive advantage. Um we're in the enviable position.
▶ 1:19:04um globally of having the the minds, the technology, the chips um and and the companies that are driving the innovation and application to deliver the valuable um results. So um you know, I just recognize like we don't want to home in on any one part of this. We need to think across the board of what it takes to maintain um American leadership um in AI and and it's an all of the above approach. That's well said. Thank you.
▶ 1:19:31Uh further Americans see and hear about AI and what it can do and they're skeptical. How and why is Cartava convinced that the positive impact on agriculture will be real and the developments utilizing AI might we see in the future that aren't already here? Yes. Yes. Thank you for the question. The um and and this has come up actually in a few other questions as well. This idea of model bias hallucination. Um I want to reiterate a couple of things.
▶ 1:20:02One is that all of the models that we produce um we rigorously validate just as we would any other discovery. We have a human in the loop. We're running assays in the lab. Um we're testing the results. It's also really important to be able to differentiate. There's so much going on in AI right now. There's an entire class of models that are focused on natural language processing. Think of all of the chat bots and the agents that are um getting wrapped around products and put into software.
▶ 1:20:30Um those are certainly capable of hallucinating and software companies are looking at how do they control those models to avoid those misbehaviors. When we're applying these models to things like biology and chemistry, it's not a matter of can it hallucinate because again we test it. It either works or it doesn't. Um and so again we're testing it to show that these models truly do work. It is not a hypothesis. Is AI going to change the way that we can innovate in agriculture? We know it to be true because we're already seeing the results and what we're able to discover. Excellent.
▶ 1:21:00Okay, Mr. Chairman, I'm uh that finishes me up, so I yield back. All right. Thank you, chairman. And again, I thank the witnesses uh for their valuable testimony today and the other members for their questions. The record will remain open for 10 days for additional comments and written questions from members. And this hearing is adjourned.