The CFS High-Level Forum on “Harnessing Artificial Intelligence, Digitalization and Data Governance for Food Security and Nutrition,” held on 30 June 2026, explored the potential and risks of artificial intelligence and digital technologies in addressing the challenges of food security and nutrition. The discussion underscored the need for governance of AI and digital technologies to ensure their responsible adoption and the effective use of data in decision-making, in order to protect and promote people-centered food systems grounded in the right to adequate food.
The CSIPM called for urgent, binding regulation of the private tech companies operating within food systems, moratoriums on high-risk technology roll-outs, and full disclosure of conflicts of interest among those involved in designing agricultural decision-support tools. Delegates also pushed back on treating data governance as a purely technical question: at its core, it is a matter of sovereignty: who controls the data, for what purpose, and under what ownership regime, are the first questions that need answering, not an afterthought.
A summary of the meeting will be presented by the CFS Chairperson at the 54th CFS Plenary Session in October, under Item II on the State of Food Security and Nutrition in the World – Strengthening Coordination and Collaborative Actions. The discussion will also contribute to identifying policy considerations for future CFS discussions or potential workstreams.
During the event, the HLPE-FSN launched its background note which explores the current evidence on the impacts of AI, the importance of data governance, the conditions needed for its responsible adoption, and the policy questions that deserve greater attention as these technologies continue to evolve. The background note was open for consultation for CFS participants, read the CSIPM input.
“We can exist without artificial intelligence. We cannot exist without what Earth provides us.” – Patti Naylor, National Family Farm Coalition (NFFC), CSIPM Data Working Group, USA
Patti Naylor was the speaker on behalf of the CSIPM at the panel on Exploring the frontiers of data and innovative AI governance needs in food security and nutrition. She warned that AI and digital technologies are being rolled out across food systems at high speed with almost no oversight of their true social and environmental costs. She called for binding governance frameworks with corporate accountability, a moratorium on new high-risk technological infrastructure until governance mechanisms are in place, and support for farmer-led alternatives like agroecology, arguing that algorithm can not replace a farmer’s real-time, generational knowledge of land, water, and pollinators.
Download the transcription in PDF
Good afternoon and thank you everyone for being here. And thank you to the CFS for bringing us this forum. We, are at a critical moment in time. Our technological advances have a potential to create an equitable, just world and make the lives of millions of people better. However, the implementation of artificial intelligence, digitalization and, and data driven technologies in the context of our food system is taking place at a very quick, pace with very little oversight or consideration of their true social and environmental costs. This is a time to question the role of these technologies in our food system rather than accepting them as inevitable. On, my flight here to Rome from the United States, someone in front of me on the airplane was watching a movie. It was Jurassic Park. That was the one where scientists decide were able to, bring back the dinosaur. It was fiction, of course, and the movie did not end well. But it m reminded me that even though we have technologies or science to develop technologies, it doesn’t always mean that we should. So, I think this is especially true when it comes to biotechnology, that is altering the basics of life itself. In the Civil society and Indigenous Peoples Mechanism our position is clear. We firmly believe that we need strong and binding governance frameworks with provisions for corporate accountability and for the dismantling of political and economic power structures of our current system. We need full assessment of digital technologies already in use. And those emerging technologies, is their use warranted or are their risks too great? We need a halt or a moratorium, as mentioned earlier today, to new technological infrastructures until governance mechanisms are in place. And we need to support, alternatives like agroecology and farmer owned and farmer led technologies that takes us beyond food security to food sovereignty and technology sovereignty.
The first question here identifies one of the major risks of artificial intelligence. But it frames it in a way that misses that, contradiction. By shaping what is seen, measured and acted upon, artificial intelligence will be choosing and actually limiting the information we receive and influencing our decision making abilities, whether those decisions are made by food producers, researchers or policymakers. As a farmer, I can tell you that artificial intelligence cannot see and know the most important aspects of food production that a farmer can see, know and understand the knowledge of the land, the wind, the water flows, the plants, animals and pollinators, as well as the pests, and in real time, as well as over seasons, over years and over generations. The question also recognizes power imbalances. We should ask, do we have the political will to address the power structures that create these unfair conditions? If we are serious about ending hunger and food insecurity. We need to name and dismantle those power structures. I would like to point out that the forum’s concept note references the strengthening collection and use of food FSN data policy recommendations. While those recommendations did have important aspects that need to be implemented, the CSIPM advocated for strong governance language in that document. But the final policy recommendations that were endorsed in 2023 contain weakened governance mechanisms as a result of the complex negotiation process. This forum is not a negotiation, Center. We are discussing things in a very, coordinated effort and I appreciate everyone’s inputs today, but we do need to think of these policy recommendations, but not as a strong basis for our governance considerations. CSIPM also agrees at the earlier points today regarding the framing of this forum and using the term harnessing, is not as relevant and impactful as it would have been if it had been governance of artificial intelligence, digitalization and data collection. I’ll finally end by saying governance needs to be binding. Needs to include not just artificial intelligence but also all of the digitalization and data technologies. It should ensure a rights based focus with accountability to social and environmental impacts and needs to include continuous reevaluation and reassessment of existing technologies, not just new ones. Thank you.
CSIPM advocates for a moratorium in the implementation of these technologies.
In the second question we’re asked how to ensure that those most affected by the risks of AI technologies are involved. We would also want to ensure that those most affected are not harmed. So, first of all, the CSIPM advocates for a pause or even a stop in the implementation of these technologies and uplift the precautionary principle. We need to name the risks, especially as we are experiencing the hyperscaling of artificial intelligence.
The Global South is especially vulnerable to these risks. But these risks are inherent and universal. They include biopiracy, land grabbing, extreme exploitation of labor, extreme exploitation of our natural environment, increased cost to the use of water, energy and materials, surveillance, loss of privacy, and questions of data ownership. The loss of Indigenous and traditional knowledge, especially for the sharing of this knowledge with future generations, is a real concern. These are serious risks to the future of humanity as these technologies and their infrastructure threaten to cross the threshold of planetary boundaries, increase the chaos of climate change and eliminate and replace labor. There’s also the concern about the enclosure of commons. Why isn’t technology considered a common good? We’re stripping away traditional rights and is a concern for data colonization.
These real and already existing risks are increasing migration, displacement, conflict and hunger. We already have rights based, proven frameworks that we need to draw on as we look to governance of artificial intelligence and digitalization. These include the indigenous people’s right to free, prior and informed consent, the UN Declaration on the Rights of Indigenous People, the UN Declaration on the rights of peasants and others working in rural areas, the care principles of indigenous data and knowledge and traditional knowledges, the collection and collective benefit, authority to control, responsibility and ethics. And there are so many others.
We don’t need to start from scratch. The benefits of artificial intelligence and digitalization in our food system are not certain. Even if we eventually have governance mechanisms in place. By placing them in this question as core infrastructures, it makes the assumption that they are neutral, necessary and inevitable. But we must first consider the alternatives, including agroecology that has a well documented benefit in regard to food security, food sovereignty, climate change and biodiversity. The future of humanity and of our precious Earth are at a crossroads. We need a vision of the future that recognizes our relationship to nature. We can exist without artificial intelligence. We cannot exist without what is Earth provides us.
“Artificial intelligence is not only a matter of technological innovation. It is fundamentally a matter of governance.” Sabrina Masinjilla, SID, CSIPM Data Working Group
Speaking on behalf of CSIPM at the panel on Exploring the way forward, Sabrina Masinjilla warned that AI and digitalization are expanding faster than the world’s capacity to govern them, concentrating control of the cloud infrastructure and algorithms behind the digital economy in the hands of a few tech giants. She joined other participants in the event calling on the CFS to establish a targeted moratorium on hyperscale AI infrastructure, and closed by questioning whether the World Food Programme’s (WFP) use of Palantir software is compatible with its humanitarian mission, given Palantir’s documented ties to military technology used in war and famine.
I am delivering this on behalf of the Civil Society and Indigenous Peoples’ Mechanism based on our analysis of the discussions today and our key requests towards the CFS.
As it has been said this morning and afternoon that Artificial intelligence and digitalisation is expanding faster that our capacity to govern it.
This is true in a context where, as highlighted by panelists and the HLPE representative, only a handful of tech giants control the clouds, infrastructures and algorithms in which the digital realm exists today, including those applications showcased here as with the potential to support climate modeling and humanitarian action.
As rights-holders we are deeply concerned about the narrative of inevitability of AI and digtialisation on food systems technology as if we are stripped of choice.
The AI industry is speculative, predatory in terms of resources and data and its economic benefits are largely unevenly distributed. As highlighted in the background note, of this meeting, there is extensive evidence of negative impacts of the AI value chain on land, water, energy, with identified risks for data ownership. At the same time, there is no scientific evidence whether generative AI can actually accrue benefits to food security and nutrition.
We therefore support the call made earlier by Mr. Pat Mooney and reiterated by Ms. Patti Naylor, but also the Special Rapporteur on the Right to Food, on the CFS to consider establishing a targeted moratorium on hyper-scale AI infrastructure because of its potential impacts on food systems, until mechanisms for independent assessments, guranteed transparency, and public accountability are effectively in place. Risk assessment is not only necessary to protect those who use the technology, but for those who bear the costs of the resource extraction and the infrastructure to materialize it.
We also need to ensure that we assess whether AI should be used -at all- in some instances, as the best tool for the job. We strongly believe that AI does not equal Innovation. Innovations that effectively support equitable and resilient food systems are grounded in grassroots experiences, and farmers’ knowledges and experimentation. They are led by famers and take place all around the world every day to effectively address farmers needs, manage risks and feed communities.
Speakers today have widely highlighted that, while a strategic resource in the digital economy, agricultural and farmers data ownership, control, and protection remain largely uncertain. In the face of no existing provisions for data governance, the deployment of data driven technologies is not urgent, rather risky.
As mentioned by Ms Patti Naylor, and the Special Rapporteur on the Right to Food, the precautionary principle is crucial and should guide the adoption of new technologies. As it is, no new medicine is authorized without rigorous clinical trials demonstrating its safety. The same logic should apply to AI systems that may affect livelihoods, human rights, and food security and nutrition.
In this context, the CFS has a crucial role to play. Besides, organising timely open and open democratic dialogues such as this, it should also embark on a process to develop a guidance for the governance of AI in food systems. This should be coupled by a mechanisms for monitoring its implementation and ensure technologies adopted in food systems are guided by the right to food, human rights, and public interest, and not agendas led by BigTech and agribusiness.
The CFS policy recommendations on Data need to be implemented specifically around:
- Risk assessments
- Protection of data and knowlegde, also as called by the Special Rapporteur on the Right to Food.
And we ask that Rome based agencies also apply these recommendations across their institutions.
We believe that the CFS is the best space to guide this process but also strengthen coherence between the CFS, FAO, and other multilateral spaces so that AI governance rules are defined within a transparent, democratic, and rights-based framework.
We would also welcome an independent report by the HLPE-FSN to prepare on the impacts of AI in food systems, particularly on employment, inequalities and injustices, and the rights of affected populations.
Artificial intelligence is not only a matter of technological innovation. It is fundamentally a matter of governance. The CFS has the responsibility to ensure that this transition is guided by human rights, public accountability, and the right to food.
Finally, we cannot ignore the role that Big Tech companies like Palantir have in the use of data and digital technologies as a weapon of war and famine, therefore WFP must reassess wether embedding the software of a complicit military contractor within their data ecosystem – pertaining the supply of food, compromises its mission.
Beyond the CSIPM statements, we highly recommend watching the full recording of the Forum, which also featured:
- Pat Mooney, International Panel of Experts on Sustainable Food Systems (IPES-Food)
- Sofía Monsalve Suárez, UN Special Rapporteur on the Right to Food
- Jim Thomas, Hopper Dean Foundation and the Philanthropic Foundations Mechanism of the CFS
- Anita Gurumurthy, Executive Director, IT for Change
- Barbara Ntambirweki, ETC Group and African Technology Assessment Platform (AfriTAP)
Join the conversation through the CSIPM Data Working Group at this link!
Without public oversight and collective data rights, the future of farming will not be governed by those who work the land, but by those who own the digital infrastructure.
– Barbara Ntambirweki, ETC Group and AfriTAP
It is imperative that the UN, the CFS, and FAO establish ex ante mechanisms to assess the impacts of these technologies, so that their use is authorized before deployment, rather than applied with no risk controls whatsoever, and no clear assignment of responsibility in the event of harm.
–Sofía Monsalve Suárez, UN Special Rapporteur on the Right to Food
We really need a moratorium on the build-out of this infrastructure, not necessarily the research, so we can assess what is happening, who is harnessing whom, and what this system is going to do to us.
– Pat Mooney, IPES-Food
There can be no food sovereignty without data sovereignty, and no data sovereignty without a legally binding right of the people to participate in data and AI governance.
– Anita Gurumurthy, Executive Director, IT for Change
The task before the CFS, which is a governance fora, is not to set out different ways this could be harnessed, is to set out how this could be governed, and how this could be governed more equitably.
– Jim Thomas, Hopper Dean Foundation