The population of the planet is going to grow from the current 7.7 billion to 9.7 billion in 2050 and nearly 11 billion by 2100, according to a 2019 report by the United Nations Population Division of the Department of Economic and Social Affairs1. These simple and cold statistics reveal two indisputable trends. On the one hand, the competition for land for different uses will add pressure to the availability of land for agriculture, while on the other a growing population inevitably implies the need for greater agricultural production for food, raw materials and energy. Add to this mix the harsh realities of global climate change due to human activity and environmental degradation, it becomes clear that extending the agricultural frontiers by further depleting existing forests does not appear a viable option2.
Intensive use of information technologies for agricultural production, known in different regions as precision farming, smart farming or digital farming, can offer a means to increase agricultural output and farmers’ profitability and to preserve the environment. While the last couple of centuries of agricultural mechanisation brought about a substantial increase in agricultural production, it also gave rise to the aggregation into larger fields of small plots each with varying soil, light and natural irrigation. To make these same fields more productive and profitable they need to be managed using methods that take into account such variations, so that each plot receives the exact amount of the inputs needed for optimal production3. This can be achieved using technology, the Internet of Things and a vast array of gadgets, all of which can be encompassed by the term “smart farming”4.
Smart farming consists of a suite of technologies rather than a single technology and according to one report its global market stood at nearly $5 billion dollars in 2016, increasing to $16 billion by 20255. The technologies deployed include: the use of smart remote environmental sensors, Artificial Intelligence, mechatronics, drones, geographic information systems, and digital veterinary applications. Smart sensors can collect vast amounts of data to forecast production outputs and anomalies. AI can be used to process that data, forecast production output and anomalies for better distribution, financial planning and mitigation. Driverless machinery can perform different tasks around the clock, with replicable precision and in the face of adverse environmental conditions. Drones are being used to gather data and control both crops and animal production. Geographic information systems allow farmers to increase production to map and project fluctuations in environmental factors while digital veterinary applications include telemedicine, trackers, wearable, monitoring and identification devices, and visual and sound recording. The use of these technologies in agricultural production invokes a range of legal issues, some of which have clear definition and others that might need present norms and interpretations to be adapted.
The impact of restrictive intellectual property legislation on food security and agricultural production has been widely discussed and deserves a series of separate articles, so in this piece I just want to highlight some of the more technology specific issues, leaving the analysis of each of them in greater detail for a later date.
The vast amount of data produced by sensors in fields and animals requires the re-tailoring of agricultural contracts to identify the different responsibilities and limitation of liabilities arising from the damaging consequences of decisions based on faulty data. At the same time, some of the data, when referring to agricultural production, environmental factors and animals, may result in the identification of the producers, which would attract the whole set of data protection laws to this data that seem to be unrelated to the use imagined by legislators. Given the sensitivity of the data collected, data security needs to be a clear legal requirement, not only at contractual level but also with regards to public safety and market transparency considerations.
There is a common, although not necessarily correct, understanding that the use of ICT in farming only leads to the processing of non-personal data, but there are an array of ways in which the collected agricultural information can be linked to an identified or identifiable persons, as defined by Article 4(1) of Regulation 2016/679. Sensors used to ensure that a specific small plot receives the exact amount of water, pesticides and fertilisers, can unwittingly collect personal data alongside the environmental conditions they are gathering: the size of the farm and its output can reveal the farmer’s income, which was clearly defined as personal data by the CJEU in Tietosuojavaltuutettu6. In the same way, technology allows very precise agricultural product traceability, as required by the General Food Law Regulation (EC) 178/2002, and necessarily includes the producer’s personal details such as the identity of the farmer and the place of origin: that is, the farmer’s address. Smart systems can operate and monitor certain aspects of farms remotely, providing food and water for animals, fertilisers and water for crops, managing energy requirements and surveilling the agricultural establishment. These systems inevitably capture information and images of the people working on the farms, or indeed their absence, all resulting in potential breaches of the right to private and personal life enshrined in Article 8 of the Council of Europe Convention on Human Rights and Article 7 of the Charter of Fundamental Rights of the European Union.
If these legal and personal data issues are not properly addressed, simply turning on some of the machinery would allow the technology providers to access a wealth of information without the farmers’ say in its use, especially as the technology can be produced in a different jurisdiction. The US case Haff Poultry, Inc. et al v. Tyson Foods, Inc. et al of 2017 exemplifies how such technology allows the sharing of inadequately anonymised data without the farmers’ consent (in that case to keep payments to growers below competitive levels) and the resultant legal implications. Even in more regulated jurisdictions, like the EU, current smart farming practices might not take into account the requirements of data protection law while the necessary interoperability of the array of systems used in smart farming risks infringement of data protection principles, such as purpose limitation, explained by the CJEU in Bara7.
Artificial intelligence in agriculture attracts all the same legal issues currently being asked of artificial intelligence in general, but there might be some specific impacts on agricultural production. The main use of algorithmic systems is to forecast different scenarios using patterns emerging from the data and, for example when management decisions lead to severe variations in agricultural output, creating surpluses that affect prices and profitability or severe shortfalls in agricultural production. The lack of transparency and accountability found in some AI systems may lead to legal lagoons and multi-party and multi-jurisdiction litigation.
The use of mechatronics in farming introduces the issue of liability for malfunction and, more importantly, undesirable production results. Some of the legal implications are well known, as are answers to questions like responsibility for damages caused by faulty machinery. However, when dealing with food, raw materials and energy production, the impact of unexpected outputs, either in quality, quantity or both, can reverberate outside the farm, bringing into question the need to redefine proximity in some tortious situations.
Drones, geographic information systems and the whole set of digital veterinary applications encapsulate such issues, requiring strong cybersecurity and regulatory compliance, and may bring into question some fundamental rights issues. For example, what are the security requirements for veterinary applications that have the potential to put unhealthy products in the consumer market through third party malign interference? What are the obligations to protect against cyber-attacks and hacking of the food system, given the impact on food security and safety? Can a farmer use technology based on drones near an airport? If not, would the airport operator or the state compensate the farmer for the potential losses, lack of profits or diminish on the value of the land?
To all of which needs to be added the fact that smart farming brings a new group of actors into the agricultural world; the technology providers with no previous links to farming. These actors are embedding complex and voluminous software licence agreements in their equipment, that end up governing both what the farmer can do with it and what happens to the resulting data.
In summary, the food, raw materials and energy requirements from agriculture imposed by a growing population, and the maintenance of a healthy environment, have the potential to be satisfied by the use of information technology in agricultural production. The technology seems to be ripe for use and the financial equation seems to favour such a move. However, given the potential influence on food security and environmental sustainability, any such developments look set to raise some legal and regulatory challenges that will have an impact on a tech law profession unaccustomed to dealing with the agri-food industry.
Dr Fernando Barrio is a Lecturer in Business Law, both in the School of Business and Management and the School of Law of Queen Mary, University of London. His research is currently around AI, accountability and Human Rights, creative industries (IP+), Smart Farming, sustainability, and the development of human-centred technology regulatory principles.
