USDA security analysts warned in May 2021 that a cyber attack could result in more chaos in the food supply chain than COVID-19. Less than a month later, JBS, the world’s largest meat processing company, was hit by a Russia-linked ransomware attack that paralyzed its factories that produce nearly a quarter of US beef and foodstuffs for other countries. JBS has paid $11 million in bitcoin to re-establish itself as an important link in the food supply chain. Then, a few months later, the hackers requested $5.9 million from the New Cooperative, an Iowa grain cooperative whose computer network runs a feeding schedule for millions of chickens, pigs and cows. Dozens of malware and ransomware attacks like these have targeted food, processor and packaging manufacturers in recent years. And last month, the FBI issued a new warning that cybercriminals are now targeting farms and food producers during critical planting and harvesting seasons.
Agriculture today faces many challenges, including climate change, labor shortages, small margins and supply chain problems caused by the war in Ukraine and COVID-19. Online data systems such as those at JBS and New Cooperative promise to alleviate some of these challenges by increasing productivity. Sensors that record agricultural data and transmit it over the Internet help farmers make decisions that can increase their returns at the lowest cost. Artificial intelligence (AI) can be used to detect patterns in agricultural data that help reduce expensive inputs like fertilizer while increasing crop yields. These advanced technologies, in turn, may help feed the more than two billion food-insecure people around the world.
However, the rapid adoption of these technologies may make computer-assisted diets vulnerable to hacking, according to a recent research paper published in The intelligence of nature’s machine. Designers of robotics and artificial intelligence in agriculture must assess risks and develop protocols for problems that can reasonably be expected. If not, their inventions may increase social and economic inequality and harm the environment.
“If the advantages of AI in agriculture are obvious, they should be. [their] Potential adversities are linked to this, said Assaf Tzakor, the paper’s lead author.
The benefits and risks of autonomous robotics and artificial intelligence in agriculture. Artificial intelligence and autonomous robots have the potential to be a boon to agriculture. AI, for example, can detect and diagnose plant diseases early, and then algorithms can guide autonomous robots to tackle problems. It can predict weather and crop yields, helping farmers to plan. It can also automate plowing, planting, fertilizing, monitoring and harvesting. In most of these tasks, it outperforms humans in processing, synthesizing and analyzing real-time agricultural data relevant to farmers’ decision making. Also, the algorithms can organize drip irrigation networks, command fleets of swarms of robots to monitor topsoil, and supervise roving weed-detection vehicles, self-driving tractors, and combine harvesters, according to Tzakor. These practices often conserve resources, save labor, reduce losses, and increase revenue.
But many of these systems are internet-based, which means hackers can try to disrupt these digital food supply chains. Cyber attacks can interfere with AI-led machines designed to harvest and inspect crops or apply fertilizers and pesticides. Such attacks may affect every link in the food supply chain, including growth, processing and distribution.
“I am intrigued by the concept of the progress trap,” Tzachor said of his motivation for research, “…the dynamics in which implementing a promising technological solution to a problem inadvertently leads to a new and more sinister problem.”
Sometimes the threat, whether unintended or intended, comes from within. For example, before a self-propelled machine can head into a field, it must be programmed by an error-prone human being. The entire crop could suffer if a program instructed an autonomous robot to use insufficient water or too much herbicide, pesticide or fertilizer. Even when the device works as intended, poor design can leave the diet vulnerable. For example, wireless sensors that detect pests and robots that use chemicals can be programmed in a way that prioritizes short-term crop yields over long-term health and safety of the environment. Mechanized tillage, which is known to damage topsoil, may exacerbate soil erosion. In short, AI’s narrow focus on agricultural crops may, in fact, ignore pest, biodiversity and pollution problems, according to the study.
Who is reaping the benefits of artificial intelligence and robotics in agriculture? Some agricultural artificial intelligence is based on data held by national and international research institutions. This data must be relevant to be useful. But those decades of institutional data focused on staple crops from rich countries such as wheat, rice and corn, according to the Tzachor study. Large data sets focusing on crops such as quinoa, cassava and sorghum produced by low-income subsistence farmers around the world are hard to find. Likewise, indigenous farmers’ sustainable approaches to cultivation, pest control and harvesting are often overlooked.
“Small farmers who grow the majority of farms around the world and feed large swaths of the so-called global south are likely to be excluded from these AI benefits,” Tzakor said. Even when relevant data exists, many small farmers live in remote areas where access to the Internet on which digital agriculture depends is poor. As a result, the increasing use of artificial intelligence in agriculture may widen the gap between commercial and subsistence farmers.
In some cases, the private sector and NGOs have partnered with disadvantaged agricultural regions to help reduce this technology gap. In sub-Saharan Africa, for example, more than 60 percent of the land is made up of small farms and 23 percent of GDP comes from agriculture. In Ghana, a private farmerline group funded the technology that underpins distribution networks. Mancius Atah, co-founder of Farmerline, said: Take Crunch. For his efforts, Atta was named the 2021 Bloomberg New Economy Catalyst.
However, even farmers with access to digital tools face problems, according to Diana Moss, president of the American Antitrust Institute. Farmers who sign technology agreements with major agricultural biotech companies often relinquish their rights to the data, resulting in what is known as a “closed cropping system.”
“The closed farming system is basically saying, ‘Look, you can only use Monsanto products. Or you can use only Dow products. “Big biotech companies are engineering their cropping systems to be non-reactive to competing technologies.”
Once farmers are locked up in a company that holds their data, Moss said, both farmers and consumers will pay higher prices: “This contributes to a very fragile agricultural supply chain.”
How can the risks of mechanized farming be reduced? Advanced technology in agriculture is not inherently bad and is poised to provide benefits. However, problems arise when humans fail to predict and prevent unintended consequences of its use.
“As in the digital sector with Facebook, Amazon and Google, it’s about having a few or one dominant player with really strong incentives to use data to control competition, to the detriment of farmers and consumers,” Moss said. “Merge control and strong antitrust enforcement are really the starting point for controlling all of this.”
CGIAR, a global research partnership between agricultural research institutes, promotes the use of the FAIR data principles that are findable, accessible, interoperable, and reusable. Farmers should be able to own and share their data at their discretion, without compromising their privacy or introducing security risks. But democratizing data in this way requires sophisticated collaboratives. Brookings has recognized efforts worthy of emulation, including Twiga Foods in Kenya and Tanihub Group in Indonesia, both of which are digitizing the small farm-to-table supply chain.
Tzachor and his team also suggest the use of “digital sandboxes” in which researchers and farmers can assess risks and make recommendations for control before systems go live. For example, HandsFree Hectare at Harper Adams University runs a hybrid electronic-physical space in which it tests unintended consequences.
Tzakor suggests that rural anthropologists, applied ecologists, ethicists, and data scientists should be invited to collaborate with computer scientists working to develop robotics and artificial intelligence for agriculture. This may act as an insurance policy against environmental damage and socio-economic disparities arising from its use.
Last month, the FBI warned the food and agricultural sector that cybercriminals are now timing their attacks on the industry to coincide with critical seasons such as planting and harvesting. Such attacks may increase the willingness of victims to pay a ransom, given that any delay could lead to an unrealized or spoiled harvest. The FBI recommended steps to prevent cyberattacks, such as backing up data, using multi-factor authentication and strong passwords, updating antivirus and antimalware software, and designing a recovery plan.
“About 50 malware and ransomware attacks have been recorded targeting food manufacturers, processors and packaging over the past two years,” Tzakor said, while acknowledging that none of them have been catastrophic so far. “This may be due to the fact that we have not yet delegated a great deal of autonomy to machines to run our farms.”