Mariana Tamari and Joana Varon, working with Coding Rights for the Tramas Project, examine how AI, automation and digital tools reshape land use in Brazil. They show how modern agribusiness combines Big Tech, Big Agro and financial interests into a highly digitalised farming model often presented as "precision agriculture." This model uses sensors, remote monitoring, automated fleets and predictive AI to promise control and efficiency.
In practice, the study finds, the approach displaces people, reduces agricultural diversity and replaces traditional, empirical knowledge with data-driven decisions. Digital systems such as rural land registries can overlay historical occupations and make collective territories invisible. When databases do not erase communities, drones and agrochemicals have been used to intimidate small farms.
The researchers call for rejecting technological neutrality, demanding transparency and decentralised governance of AI systems. They urge public infrastructures that enable collective participation and the protection of social territories, and they value regenerative, agro-ecological and ancestral practices.
Difficult words
- agribusiness — Large businesses involved in industrial farming
- precision agriculture — Farming using sensors and technology for accuracy
- displace — To move people away from their usual landdisplaces
- diversity — Different kinds of crops or farming methods
- empirical — Based on observation and experience, not theory
- data-driven — Based on analysis of collected digital information
- transparency — Open sharing of information and clear processes
- decentralised governance — Decision-making shared among many local organisations
Tip: hover, focus or tap highlighted words in the article to see quick definitions while you read or listen.
Discussion questions
- How might replacing empirical knowledge with data-driven decisions affect small farmers?
- What kinds of public infrastructure could help communities take part in decisions about land and AI?
- Do you think precision agriculture can be changed to protect local communities? Why or why not?
Related articles
Reducing unsafe responses in large language models
Researchers studied how large language models (LLMs) handle safety and tested training methods to reduce unsafe outputs while keeping performance. They identified key challenges and a technique that preserves safety during fine-tuning.
New commercial bee feed improves winter survival
A study published in Insects found that a new, nutritionally complete commercial feed helped honey bee colonies stay healthier through two winters and the California almond pollination, reducing winter losses and raising hive productivity.