Happy Monday ranchers,

Peter Thiel’s Founders Fund led a $220 million investment in Halter, the New Zealand-based virtual fencing company, at a $2 billion valuation. It was one of the largest agtech raises in history, and every major ag outlet covered it the same way: Silicon Valley finally takes cattle seriously.

That framing isn’t wrong. It’s just incomplete. Because the same week that round closed, a Ph.D. student from Wheatland, Wyoming published a 92%-accurate AI model that identifies subclinical heart failure in cattle from a photograph. He built it with 7,200 images, a cell phone camera, and a partnership with his university’s computing department. No Founders Fund. No $2 billion valuation. Just a rancher’s kid who grew up wiping blood off hearts on a slaughter floor and thought there had to be a better way.

Both stories are real technology solving real problems. But they raise different questions — and the difference matters. One is a venture-backed platform built to generate returns for investors, with producers as the customers and their herd’s behavioral data as a quietly accumulating asset. The other is a land-grant research project with the explicit goal of improving producer margins.

This week we unpack both. In Simple Terms covers the Wyoming heart model — what brisket disease is actually costing you, why it’s been invisible until now, and what this tool could change for feedlot operators and high-altitude ranchers. The Deep Dive takes Halter’s $220 million raise seriously as a business story, because the question buried inside it — who owns the value your herd’s behavior creates — is one every producer using connected livestock technology should be asking.

First time reading? Sign up for free here

King Ranch Uses Virtual Fence Tech to Help Ranchers Navigate Potential Screwworm Outbreak | Texas Public Radio
The King Ranch Institute is deploying virtual fencing technology to help cattle producers contain and manage herds in response to the threat of New World screwworm — a real-world disease biosecurity application that goes well beyond the typical grazing management use case.

Halter’s ‘Cow-gorithm’ and Solar-Powered Collars Are Changing the Virtual Fencing Landscape | RFD-TV
RFD-TV's Elena Chavez walks through how Halter's AI-driven "Cow-gorithm" works in practice, with field visits to hundreds of Colorado ranches confirming strong demand and early adopters praising the system's ability to simplify herd management.

How Artificial Intelligence Is Useful to Cattle Producers | Farm Progress
Farm Progress distills the practical on-ranch applications of AI — from health monitoring to feed efficiency — into a producer-facing guide, providing a useful counterbalance to all the hype-heavy coverage currently surrounding the space.

Beyond the Sale Barn: How AtTheYards Is Digitizing Cattle Marketing | Drovers
Drovers covers AtTheYards, a platform bringing cattle marketing online with digital listings, real-time bidding, and transparent pricing tools — a direct challenge to the opacity of traditional sale barn commerce that producers of all sizes should be watching.

AI or ET: Which Is Best for Your Cattle Operation? | Farm Progress
Farm Progress tackles the practical question of when to use artificial insemination vs. embryo transfer — a timely comparison given the wave of reproductive tech investment happening across the industry right now.

Halter Raises $220M Series E to Accelerate Global Expansion of Virtual Fencing | Ventureburn
Halter closed a $220 million Series E led by Founders Fund — one of the largest capital raises in agtech history — at a $2 billion valuation, with one million solar-powered collars now sold and 60,000 miles of virtual fencing built by U.S. ranchers since the company's 2024 launch. The same investors behind SpaceX, Palantir, and Stripe are now betting on cattle collars.

Halter Says It's Not an Agtech Company on the Heels of Its $220M Series E | AgFunder News
AgFunder's deep-dive raises the key question: at a disclosed collar price of $5–8, hardware sales alone don't explain a $2 billion valuation — the real story is whether Halter's software layer can generate the recurring, mission-critical adoption that turns a clever piece of hardware into an industry standard. Essential reading for anyone trying to understand the Halter bet.

Scanabull Raises NZD $1.1 Million for Cattle Weighing Tech | ChannelLife
Scanabull, a New Zealand agritech startup, raised NZD $1.1 million in a round led by Sprout Agritech, with support from Enterprise Angels and Callaghan Innovation's Deep Tech fund, to commercialize its AI-powered 3D smartphone scanning technology that estimates cattle weight without a physical scale.

The 'Cowgorithm' Is Here: How Peter Thiel's Latest $2 Billion Bet Is Changing Farming Forever | Inc. Magazine
Inc. Magazine highlights that Halter's proprietary Cowgorithm is trained on seven billion hours of animal behavior data and that U.S. ranchers have built 60,000 miles of virtual fencing since the company's 2024 launch — a mainstream business media take that signals virtual fencing has officially crossed into cultural visibility beyond ag trade press.

Integrating Agricultural Development and Methane Mitigation as Part of the EU Livestock Strategy | Clean Air Task Force
The Clean Air Task Force published a policy paper calling on the EU to weave methane reduction targets directly into its livestock development strategy — a preview of the regulatory and funding frameworks that are likely coming to the U.S. beef sector in the years ahead.

How 3D Imagery and AI Technology Weigh Large Animals | SunLive
Bay of Plenty's SunLive profiles Scanabull's approach to contactless cattle weighing using iPhone LiDAR and AI vision — a technology that could eventually eliminate the stress, labor, and infrastructure cost of traditional squeeze-chute weighing at scale.

Assam Records First IVF-Born Calf Using Embryo Transfer Technology | India Today
India's northeastern state of Assam has produced its first IVF calf through embryo transfer technology, marking another data point in the global spread of advanced reproductive tools beyond research centers and into regional livestock programs.

IVRI Achieves Breakthrough in Sahiwal Calf Production | The Usthadian
India's premier veterinary research institute is successfully producing calves of the Sahiwal breed — prized for heat tolerance and disease resistance — through advanced reproductive technologies, signaling that climate-adapted cattle genetics are increasingly being pursued through lab-assisted methods.

From Pasture to ROI: Implementing Autonomous Systems in Livestock and Pasture Management | GreenProphet
GreenProphet examines the real-world economics of deploying autonomous systems — drones, sensors, virtual fencing — across livestock and pasture management, with a focus on the ROI math that will ultimately determine adoption timelines across the industry.

Gov. Gianforte Highlights Innovation in Agriculture | Tri-State Livestock News
The Tri-State Livestock News covers Montana Governor Gianforte's public push for agricultural innovation — a signal that state-level policy support for ag tech is gaining political traction in beef-producing states, which could translate into grants, pilot programs, and faster regulatory pathways for new tools.

A Wyoming grad student built an AI that detects subclinical heart failure in cattle from a photograph. Here’s why it matters commercially — especially if you run cattle at elevation.

Brisket disease is the common name for bovine high-altitude pulmonary hypertension — a condition where elevated pulmonary artery pressure causes the right side of the heart to enlarge and eventually fail. It’s been a known cost of ranching above 8,000 feet for generations. What hasn’t been well understood is how widely its effects extend below the death-loss threshold.

Most producers think of congestive heart failure as a death loss problem. At roughly 0.15% of all fed cattle, the death numbers look manageable. But death loss is only the surface. Unpublished research tied to this project shows that between 4 and 7 percent of all fed cattle carry severe subclinical heart failure — animals scored 4 or 5 on the standard 1-through-5 cardiac evaluation scale. These animals don’t die. They don’t get pulled. They quietly underperform: about 40 pounds lighter on live weight and roughly 10 cents higher on cost of gain. At $235 live, applied across 4 to 7 percent of a pen, that drag is substantial. It’s just been essentially invisible.

Because detecting it required a trained evaluator on a slaughter floor, manually scoring each heart as it came off the line. It’s slow, it’s unpleasant, and it doesn’t scale. The result is that most of the industry has been managing a cost it couldn’t see, let alone quantify.

Chase Markel, a Ph.D. student from Wheatland, grew up ranching at altitude and watched brisket disease take cattle his family could never fully predict or prevent. He spent years collecting cardiac data the traditional way — on slaughter floors, by hand — until he crossed paths with an anthropologist in the University of Wyoming’s School of Computing who had built AI models to identify ancient structures from satellite imagery.

The insight was straightforward: if you can teach a computer to recognize ruins from space, you can teach it to score a heart from a photograph. Markel didn’t know what an image classification model was when he started. Working with UW’s computing faculty, he built one — training it on 7,200 heart images scored against Dr. Tim Holt’s established 1-through-5 cardiac evaluation system. The result is 92% accurate and runs from a standard cell phone camera. Instead of a specialist scoring each heart by hand, a camera fixed over an offal line can capture images the model scores instantly, linking cardiac condition directly to carcass performance data.

For feedlot operators, the immediate application is at the processing plant: connecting heart scores to hot carcass weight, yield grade, and marbling data at scale for the first time. That connection starts to answer questions the industry has been guessing at for decades — which animals are dragging pen performance, and whether subclinical disease is why.

For high-altitude cow-calf producers across Colorado, Wyoming, Montana, and the northern Rockies, the longer-term promise is a selection and management tool. Brisket disease has always been part of the cost of doing business at elevation. Markel is building the beginning of a way to quantify that cost, breed around it, and eventually identify at-risk animals before they ever reach a feedlot.

Halter just raised $220 million at a $2 billion valuation. The collar is useful. But the asset Founders Fund is actually investing in is your herd’s behavioral data — and producers should understand that before they collar up.

The headline is straightforward: Founders Fund, Peter Thiel’s venture firm, led a $220 million Series E round in Halter, the New Zealand-based virtual fencing company, valuing it at $2 billion. The round closed oversubscribed, meaning more investors wanted in than Halter had room for. It’s one of the largest agtech raises in history.

The coverage has been uniformly positive, and not without reason. Virtual fencing solves a real problem. Halter now has more than one million collars deployed across 2,000-plus operations in New Zealand, Australia, and the United States. Since launching stateside in 2024, American ranchers using the system have built over 60,000 miles of virtual fence lines. The labor savings are real. The grazing management flexibility is real.

But the framing of this deal as “Silicon Valley finally takes cattle seriously” misses the more important question. What, exactly, is a $2 billion venture capital bet on a fencing company actually about? The answer tells producers something important about the technology relationship they’re entering.

GPS hardware, solar panels, Bluetooth connectivity — none of it is proprietary. These are components. What Halter has that no competitor can replicate quickly is the Cowgorithm: the machine learning model trained on the behavioral data generated by more than one million animals, every day, across three countries and widely varied terrain.

Every time a cow responds to a boundary cue, ignores a vibration, shifts her grazing pattern, changes her rumination rate, or flags an estrus cycle, that data point feeds into Halter’s central model. The model gets smarter. The platform gets more valuable. The rancher who signs up next year gets a better product — built partly on patterns your herd taught the system this year.

Industry analysts have noted that Halter’s scale advantage — with over 100,000 collars generating training data — gives it a dataset that competitors like Vence and eShepherd cannot currently match. That lead matters because accuracy in variable terrain is the core technical problem in virtual fencing. The more real-world behavioral data you have from cattle in mountains, canyon country, and open plains, the better your boundary algorithms perform. This is what venture capitalists call a data moat. The more data flows through the system, the wider it gets, and the harder any competitor finds it to close the gap.

Here’s the analogy that clarifies the relationship. Think of Halter’s collars like a leased bull. You’re paying for the service. But the calves — in this case, the behavioral data your herd generates — go home with someone else.

Craig Piggott, Halter’s founder, built spacecraft at Rocket Lab before deciding that farming was a bigger unsolved problem. He told AgFunder this week that he has never viewed Halter strictly through the agtech lens. Blackbird, one of the round’s investors, described the company’s mission as making “half the planet’s habitable landmass more productive and sustainable.”

That language doesn’t describe a fencing company. It describes a data platform that happens to work with cattle — for now. The subscription model — $5 to $8 per animal per month — is what investors love because it’s recurring and grows with every collar. But the strategic asset accumulating beneath that subscription revenue is a global behavioral dataset for livestock that no one else is close to building at this scale.

Peter Thiel built Palantir, one of the most powerful behavioral data analytics companies in existence. He understands what large-scale behavioral datasets are worth better than almost anyone alive. If he’s willing to put $220 million behind the behavioral patterns of cattle, it’s worth asking what that data is worth to the people whose animals are generating it.

This is not a new pattern. John Deere paid $305 million for Blue River Technology in 2017, ostensibly for the see-and-spray precision herbicide system. The strategic purchase was the dataset: millions of images of crop rows in real-world conditions, continuously expanding with every acre sprayed. That dataset now underpins Deere’s broader autonomy platform. The hardware was the data collection mechanism. The data was the asset.

Google’s search dominance is the same structure. The more searches run through the system, the better the results get. Better results mean more searches. More searches mean better results. The flywheel compounds. By the time a competitor has enough query data to match Google’s accuracy, the gap has already widened.

Halter’s data moat works the same way. More collars generate more behavioral data. More behavioral data makes the Cowgorithm more accurate. Greater accuracy wins more customers in harder terrain. More customers in harder terrain generate more diverse behavioral data. The cycle is self-reinforcing — and at a million collars, it’s already running.

None of this makes Halter a bad tool. The labor savings are real, and for ranchers managing large grazing rotations with a shrinking workforce, virtual fencing solves a genuine problem. But the terms of the relationship between a $2 billion platform and its producer-customers deserve more scrutiny than they’re currently getting.

The technology is real. The labor savings are real. But so is the data, and somebody is going to own it. The question is whether producers get a seat at that table — or just a monthly invoice.

Put both this week’s stories next to each other and the contrast is clarifying. Halter’s model creates value that flows primarily to Halter and its investors, with useful tools for producers along the way. Markel’s model is being built at a land-grant university with the explicit goal of improving producer margins — the institution type that has historically produced the most producer-aligned agricultural technology in American history.

Neither is the whole future of AI in beef. But together they frame the question every producer should be asking as these tools proliferate: Is this technology working for me, or am I working for it?

That’s not an argument against adopting virtual fencing or any other connected livestock tool. It’s an argument for reading the data terms as carefully as you read the subscription price.

On the Halter round: the next signal to watch is whether Halter announces any formal data governance commitments for producers — portability rights, export tools, or third-party auditing of how herd behavioral data is used. At a $2 billion valuation heading toward a likely IPO, pressure from producers and breed associations to clarify those terms is both reasonable and timely. We’ll cover any developments.

On Chase Markel’s heart model: watch for the full patent filing later this year and any announced partnerships with processing plants or breed associations. The liver abscess model he’s developing in parallel is the one to watch for feedlot operators specifically — it targets a cost that is well understood but almost entirely unquantifiable in live animals today.

Also this week: we skipped the 11th Circuit’s ruling on Florida’s lab-grown meat ban and the ongoing Argentina import situation — both deserve more space than we had. The lab-grown ruling in particular has legal nuance that the celebration in conventional beef circles is glossing over.

BeefTech.News – Keeping you ahead of the herd.

One story cost $220 million. The other cost a lot of hours on a slaughter floor. Both are worth understanding. Forward this to a producer who should be paying attention to both.