How America’s Largest Meat Company Leverages Palantir’s Surveillance Tech

The Trump Administration’s warp-speed restructuring of the federal government, mass integration of government data and vast tracking of immigrants would likely not be possible without a key technological partner: Palantir. The software company provides the digital architecture powering what former employees describe as an authoritarian agenda. Yet the company’s staggering commercial sector growth — up by 121 percent over the past 12 months — has attracted far less scrutiny. America’s largest meat company, Tyson Foods, became an early commercial adopter of the software in 2020, establishing a blueprint for how large food corporations can leverage Palantir’s surveillance technology.

Palantir advertises its partnership with Tyson as a flagship example of how its software is “transforming food and beverage businesses for the AI era.” Tyson uses Palantir’s Foundry platform, which has been described by a former employee as a “super-charged filing cabinet” for its unique capacity to digest endless amounts of data. It’s easy to see how this data-mining technology could be incredibly useful in managing complex food supply chains with many variables — and indeed, Palantir claims that it saved Tyson $200 million over the course of two years.

Yet Tyson’s deployment of Palantir also showcases the limits of the technology in protecting a vulnerable workforce. The company used Palantir to forecast Covid-19 infections among its meatpacking workers, according to Scott Spradley, Tyson’s former Chief of Technology. This data was then used to prepare for supply chain disruptions as workers fell predictably sick, while Tyson extensively lobbied against mandated worker protections.

Tyson Foods did not respond to a request for comment by the time of publication.

Palantir’s Expansion into the Food Industry

The U.S. food industry’s workforce is particularly vulnerable to surveillance technologies. An estimated 2.1 million immigrants work across the entire U.S. food supply chain, from farmworkers to food processing workers to grocery store workers — a workforce that has long been subject to heightened surveillance to further immigrant control and raids, which have significantly expanded under the Trump Administration. And this heightened risk of deportation  makes food chain workers — fearful of speaking  out because of their immigration status — more vulnerable to corporate abuse.

While there is no indication that Palantir is currently being used by food companies to monitor employees, food corporations have been investing in other AI technologies to this end and Palantir has facilitated the internal surveillance of employees in other industries.

For instance, Tyson and JBS, the world’s two largest meat companies, have invested in smart watches used to monitor workers’ movements. In a press release, a JBS spokesperson describes this technology as a “wearables and analytics platform” that will “provide us insight into how each employee responds to ergonomic and process changes by digitizing individual worker motion.” This raw data is fed to an AI algorithm, which converts it to metrics displayed on a dashboard for supervisors to monitor.

Palantir is not part of these smart watch projects, and it has repeatedly stated that it does not harvest any raw data. Instead, Palantir is able to quickly integrate existing data already stored by the company without making any changes to the underlying datasets. Yet as Tyson’s venture capitalist arm invests in smart watches and other forms of AI, it’s easy to see how the mass integration of data across varying data fields could lead to privacy concerns. Experts have warned about the compounding risks to privacy “that arise when huge datasets are merged and analyzed, which is the premise behind Palantir’s business model.”

Tyson first contracted with Palantir in 2020, as meatpacking plants emerged as the epicenter of the pandemic, to predict infections among its poultry plant workers down to nearly an exact figure. “In a very short amount of time we started working on figuring out how bad the spread is going to be, which plants are going to be affected, how quick are they going to be affected,” said Tyson’s Scott Spradley in a keynote speech at a Palantir-sponsored conference.

“I remember telling our CEO, “Hey, this plant in seven weeks is going to have 880 people infected and we’re going to shut that plant down,” said Spradley. “So that day hit. 883 people were infected because we were driving prediction with Palantir.”

Though Tyson had the technology to forecast infections, the company didn’t use this data to heighten protections among workers in advance of projected outbreaks at its meatpacking plants, at least none that were made public. Instead, this software was largely used to protect Tyson, ensuring that infections among workers didn’t disrupt the company’s intricate supply chain. By forecasting worker infections, Tyson could plan in advance for irregular inventories from plant closures and labor shortages, as Spradley explained: “This plant’s going to go down, so we need to start shifting product to this place.”

In Spradley’s speech, the only protective measure for workers mentioned was Tyson’s vaccine mandate, which relied on Palantir to verify vaccination records.

It was later revealed, in a 2022 House investigative report, that Tyson and other meat producers lobbied the first Trump Administration to shield the industry from liability for worker deaths and illnesses and exempt the industry from the stay-at-home mandate. The House Select Subcommittee on the Coronavirus Crisis determined that Tyson and other major meatpacking companies failed to take adequate coronavirus precautions, “despite awareness of the high risks of coronavirus spread in their plants.” Tyson possessed especially detailed awareness of the risk that Covid-19 posed to its workers, raising ethical questions about the responsibility that companies have to act on the data availed by Palantir.

Tyson is part of a growing wave of major food corporations — including General Mills, Wendy’s, Beyond Meat and Aramark — to embed Palantir in their operations. The U.N. World Food Program, the world’s largest humanitarian food aid program, is another adopter of Palantir. As the software giant appears poised for further growth in the food sector, supported by energy-intensive data centers being rapidly built across the country, it’s worth asking, sooner rather than later: what are the risks of a food system that runs on Palantir?

Palantir’s Ethical and Human Rights Concerns

Like many emerging forms of AI-driven technology, Palantir is malleable and responsive, designed to reflect the motives of its users, whether that is generating more profits or improving efficiencies or protecting its workers. Named after the ‘seeing stones’ in Lord of the Rings, Palantir can give companies and governments an unusual level of insight, but this doesn’t guarantee this information will be harnessed for good.

And in the wrong hands, like the seeing stones, the integration of vast pools of data has the risk of being weaponized for unethical or even illegal ends.

These risks are something that Palantir’s employees acknowledge, to an extent, even before being officially hired. In the final step of the hiring process, prospective employees typically undergo what is known as the “founder interview,” meeting with one of the company’s four founders, according to a former Palantir employee, who asked for anonymity due to a lifelong non-disparagement agreement he signed with the company.

In this final interview stage, “they would ask you a disarming question to catch you off guard, then try to read your soul to see if you were going to put the mission at risk or not,” the former employee told Sentient over Signal. “An example of a question a lot of people got (mine was similar to this) was: ‘How would you feel if software you wrote resulted in people dying due to a bug?’ Follow up: ‘What if this wasn’t a bug, but the software’s actual purpose?”’

“They made no excuses that Palantir’s software was powerful and would be used to ‘get the bad guys’  and ‘protect Western democracy,’ whatever that meant,” adds the former employee.

By now, it has become clear that these questions were not just far-flung hypotheticals. Palantir’s contracts with military and law enforcement agencies have demonstrated that the company is a willing partner to institutions that, whether by design or accident, kill people. Its 2024 contract with the Israeli military has especially provoked outrage, prompting accusations of enabling Israel to commit a genocide. Palantir also recently inked a $10 billion contract with the U.S. Army, building a “comprehensive framework” for the military’s data. And Palantir is behind a new $30 million mass surveillance system, deployed by Immigration and Customs Enforcement (ICE), to track and deport unauthorized immigrants.

There are mechanisms built into Palantir’s design to limit the company’s clients from accessing sensitive data, according to the former employee. For example, there are safety controls to prevent employees in intelligence agencies from accessing information that they don’t have clearance to see, with an option to flag someone who does have the permission to view this data in order to share it. There are also audit logs, owned by Palantir’s clients, that can monitor what information individual employees access in the system.

“The problem is all the safeguards in the world don’t mean anything if nobody uses this feature,” adds the former employee, who never anticipated this system would be abused when he was initially hired. “You still have to trust that your government will follow their own laws, another modern-day situation that back then seemed unimaginable.”

And these built-in measures have repeatedly proven insufficient. In July, Palantir was questioned by members of Congress about likely violating multiple federal laws by merging across federal agencies to create a searchable “database” of Americans, raising concerns about how this software is enabling the mass surveillance of both unauthorized immigrants and U.S. residents. The U.S. government relies on Palantir’s Foundry platform to compile this potentially illegal database — the same platform that Tyson deploys.

As it stands now, it’s not clear how most food companies are utilizing Palantir beyond the often vague, public-facing statements. But as Palantir further expands into the commercial sector, so do concerns about using the software to compile sensitive data.

While it may be common for defense technology to be wrapped in opacity, arguably for the sake of protecting national security, Palantir maintains a similar level of secrecy even in non-military contexts, Ilia Siatitsa, a program director and senior legal official at London-based nonprofit Privacy International, tells Sentient. She sees this as a problem.

“We have to just rely on promises from a company, but how is that good enough? How is that sufficient? Companies have never lied before? That’s what I find, from a public policy perspective, problematic,” says Siatitsa.

“Once [Palantir] moves the same product to the food industry, for instance, or the health sector, or all these other public sectors, then I think the citizens have a right to know a bit more about what exactly the product is doing, how it’s used and what exactly they aggregate and for what purpose,” says Siatitsa.

Siatitsa points to Palantir’s partnership with the U.N. World Food Program, the world’s largest humanitarian organization. When this partnership was announced in 2019, Privacy International and other nonprofits wrote an open letter to David Beasley, the executive director of the U.N. World Food Program, warning that “the partnership has the potential to seriously undermine the rights of 90 million people the WFP serves.”

“Nothing has been transparently shared about the procurement process that WFP set up to engage Palantir,” states that 2019 letter. “Given the gravity of these concerns, building in transparent checks and balances — such as third-party audits, open procurement and contract or agreement transparency — seems essential.” But this letter didn’t result in any greater levels of disclosure. Even today, “we have no information how this collaboration is going,” says Siatitsa.

As for Tyson, it remains unclear how the company is currently utilizing Palantir. In his 2022 remarks, Scott Spradley appeared to be brimming with excitement about all the ways Palantir can be further integrated into Tyson’s operations. “And we’re just continuing to find more and more, once you get all your data into Palantir, into Foundry, your kind of use case exploitation is unlimited,” said Spradley. “There’s nothing that you really can’t…we haven’t found a zero sum game there yet.” Of course, there are legal limits to how companies and governments can deploy Palantir, but that is increasingly appearing irrelevant.