Second Meeting of Global Initiative on AI for Agri-Food Systems (GI-AI4FS) held at IMC 2026 in New Delhi

1. At a Glance

• GI-AI4FS is a UN Global Initiative. The International Telecommunication Union (ITU) coordinates it, working closely with FAO, IFAD and WFP. Its purpose is to build global cooperation on using AI and IoT in agriculture and food systems. [1]
• India hosted its second meeting alongside India Mobile Congress (IMC) 2026 in New Delhi. The hosts were TEC (DoT, Ministry of Communications) and ICAR (Ministry of Agriculture & Farmers Welfare). [1]
• Why it matters for UPSC: the topic brings together digital public infrastructure, AI governance, food security and international standard-setting. It fits GS-II (international institutions) and GS-III (agriculture technology, science and technology). [1][2]
• The initiative's aim is trustworthy, interoperable, farmer-centric AI and international cooperation on governance and standards. [1]

2. Why in the News

• 10 Oct 2026: PIB reported that the second GI-AI4FS meeting was held alongside IMC 2026 in New Delhi. [1]
• Before the meeting, a 2-day International Hackathon on "AI for Agri-Food Systems" was held on 8–9 Oct 2026 at the ICAR campus, New Delhi. TEC and ICAR hosted it jointly. [1]
• The meeting worked on the initiative's governance architecture: Working Group Chairs and Co-Chairs, Terms of Reference and priority work streams. [1]

3. Background & Evolution

• Origin: FAO says it set up the initiative with ITU and other partners as a multistakeholder initiative. [2]
• June 2026: the first meeting was held at FAO Headquarters. [2]
• 6–10 July 2026, Geneva ("Triple Week"): FAO pushed agrifood systems up the AI agenda at the WSIS Forum, the AI for Good Global Summit and the first-ever Global Dialogue on AI Governance. [2]
• 8–9 Oct 2026: the International Hackathon on AI for Agri-Food Systems was held at the ICAR campus, New Delhi. [1]
• Oct 2026: the second meeting was held alongside IMC 2026 in New Delhi. [1]
• Related Indian initiatives:
• Bharat-VISTAAR, an AI-based scheme for farmers, was launched in Jaipur by Union Minister Shivraj Singh Chouhan while the India-AI Impact Summit 2026 was under way. [3]
• At the same launch, a farmer-centric AI Hackathon and the AgriKosh AI strategy roadmap were announced. [3]
• The Summit held a session on "AI for Inclusive and Resilient Agricultural Food Systems". [4]

4. Core Static Facts

Item	Fact
Full name	Global Initiative on AI for Agri-Food Systems (GI-AI4FS) [1]
Nature	UN Global Initiative [1]
Coordinator	ITU [1]
Partner UN bodies	FAO, IFAD, WFP [1]
Chair	Dr. Sebastian Bosse [1]
First meeting	June 2026, FAO Headquarters [2]
Second meeting	Alongside IMC 2026, New Delhi; PIB release dated 10 Oct 2026 [1]
Indian host (lead)	Telecommunication Engineering Centre (TEC), DoT, Ministry of Communications [1]
Indian co-host	ICAR, Ministry of Agriculture & Farmers Welfare [1]
Participation	About 70 delegates in person and 25 online [1]
Participant types	International organisations, government departments, agricultural research institutions, academia, industry, standards development organisations [1]
Keynote speakers	Krishnan Pallassana (FAO); Dr. Bilel Jamoussi (Deputy Director, TSB, ITU); Dr. A.K. Nayak (DDG–NRM, ICAR, DARE); Syed Tausif Abbas (Sr. DDG & Head, TEC) [1]
Governance agenda	Governance and coordination structure; Working Group Chairs, Co-Chairs and members; Working Group Terms of Reference; priority areas and initial work streams [1]
Technologies named	Connected sensors, weather stations, drones, satellite systems, AI analytics [1]
Statutory basis	None. It is a voluntary multilateral initiative, and no Act or Article applies. [1]

5. Multi-Dimensional Analysis

Economic

• Precision agriculture and better use of resources can cut input costs. [1]
• Better traceability and lower post-harvest losses make supply chains more efficient. [1]

Scientific / Technological

• The meeting combined IoT (sensors, weather stations, drones) with satellite systems and AI analytics. Uses include early detection of crop diseases and decision support for farmers. [1]
• Interoperability and standards are central to the initiative. ITU's Standardization Bureau (TSB) took part, which shows the standards-setting angle. [1]

Geopolitical / Strategic

• India hosted a UN-system initiative only months after its first meeting. That strengthens India's image as a convenor on AI for the Global South, following the India-AI Impact Summit 2026. [1][2][3]
• The initiative brings telecom (ITU) and food bodies (FAO, IFAD, WFP) together, an unusual pairing across the UN system. [1]

Ethical / Governance

• The meeting emphasised responsible, trustworthy and farmer-centric AI. [1]
• FAO's framing: "No Farmer Left Behind is not just an aspiration. It is also a reality test." (Prof. Charles Spillane, FAO Chief Scientist). The concern is that AI benefits could be captured by large farms only. [2]

Administrative

• A telecom body (TEC/DoT) and an agriculture research body (ICAR/DARE) co-hosted, which is a model of inter-ministerial convergence. [1]
• The hackathon was a way to turn innovation into practical applications. [1]

Social

• Farmer-centric design and multilingual advisory services (for example, Bharat-VISTAAR, which draws on AgriStack, ICAR, IMD and state systems) can help smallholder farmers be included. [1][3]

6. Recent Developments (last 12–18 months)

• Feb 2026: the India-AI Impact Summit 2026 held a session on AI for Inclusive and Resilient Agricultural Food Systems. [4]
• Feb 2026: Bharat-VISTAAR was launched in Jaipur, with an AI Hackathon and the AgriKosh AI strategy roadmap announced alongside. [3]
• June 2026: first GI-AI4FS meeting at FAO Headquarters. [2]
• 6–10 July 2026: Geneva Triple Week (WSIS Forum, AI for Good Summit, first Global Dialogue on AI Governance), where FAO pushed the agrifood AI agenda. [2]
• 8–9 Oct 2026: International Hackathon on AI for Agri-Food Systems at the ICAR campus, New Delhi. [1]
• Oct 2026: second GI-AI4FS meeting alongside IMC 2026, which included a hackathon showcase and award ceremony. [1]

7. Prelims Hooks

• The GI-AI4FS is coordinated by ITU, not FAO. [1]
• Its UN partners are FAO, IFAD and WFP. [1]
• Its Chair is Dr. Sebastian Bosse. [1]
• The first meeting was held at FAO Headquarters in June 2026. [2]
• The second meeting was held alongside India Mobile Congress 2026, New Delhi. [1]
• The Indian lead host was TEC, which sits under DoT, Ministry of Communications, not MeitY. [1]
• The co-host was ICAR, under DARE, Ministry of Agriculture & Farmers Welfare. [1]
• Attendance was about 70 in person and 25 online. [1]
• The preceding hackathon ran on 8–9 Oct 2026 at the ICAR campus, New Delhi. [1]
• ITU was represented by the Deputy Director of the Telecommunication Standardization Bureau (TSB). [1]
• The meeting's focus was the application of IoT and AI in agri-food systems. [1]
• The first-ever Global Dialogue on AI Governance was held during Geneva's Triple Week in July 2026. [2]
• Bharat-VISTAAR draws on AgriStack, ICAR, IMD and state systems for multilingual farm advisories. [3]

8. Who the 'Farmer-Centric' Promise Can Leave Out

• India's farm AI depends on records of who owns land
• India's digital farm system, AgriStack, gives each farmer a 'Farmer ID' (a digital identity, like Aadhaar for farmers) [5].
• This ID is made by linking a farmer's details to digitised land records (land ownership papers kept on computer) [5].

• More than 10.31 crore Farmer IDs had been made by 3 August 2026 [5].

• Farmers who till land but do not own it may stay invisible

• Tenant farmers and sharecroppers usually do not have their name on the land record. Many women who farm do not either.
• If advisory tools such as Bharat-VISTAAR draw on AgriStack [3], these farmers may get weaker advice, or none.

• FAO's Chief Scientist called 'No Farmer Left Behind' a 'reality test' [2]. In India, this is where that test will be failed or passed.

• What 'farmer-centric' should mean in practice

• It is not enough to design the app well. The farmer first has to be in the database.
• Any AI service built on Farmer IDs should report how many tenant and landless farmers it actually reaches.

9. A Sensor Is Useless Without a Signal

• The tools in the note all need a network
• Sensors, weather stations and drones send their data over a connection. AI analytics then sends advice back [1].

• Where there is no network, none of this works.

• Villages are far less connected than cities

• FAO data for developing countries (2020): 65% of city people had internet access, but only 28% of rural people [7].

• The World Bank finds that many smallholders (farmers with very small plots) cannot afford smartphones or data plans. Low digital skills and unreliable electricity also hold them back [8].

• This is why a telecom body leading the work matters

• TEC sits under the Department of Telecommunications (DoT) [1]. DoT also plans rural networks.
• The real test of telecom–agriculture convergence is whether villages get network coverage. Holding a meeting at IMC does not prove it.

10. Who Owns the Data a Farm Produces

• AI needs farm data, and that data has value

• Soil readings, crop images and yield records train AI models. Whoever holds the data can build and sell services on it.

• India's stated safeguard is consent

• The Agriculture Ministry says AgriStack follows the Digital Personal Data Protection (DPDP) Act, 2023 and the DPDP Rules, 2025 [5].
• Farmer data is to be collected only with the farmer's consent. The farmer controls who it is shared with, and for what purpose [5].

• The State Farmer Registry is federated: each state keeps ownership of its own farmers' data, instead of one central database [5].

• Why consent alone may not protect a small farmer

• A farmer with little reading ability may tick 'yes' to get a scheme benefit without understanding the terms.
• Global standards set at ITU decide how data flows between systems. If they are built only for easy sharing, consent rules at home become harder to enforce.
• So India has a stake in what the Working Groups decide on data, not only on technology.

11. Is a Voluntary Global Forum Worth India's Effort?

• The strongest objection
• GI-AI4FS has no law behind it and no binding rules [1].
• So far, its second meeting only set up its own structure: Working Group chairs, Terms of Reference and work streams [1].

• It has no fund of its own. Meanwhile, the basic gaps are local ones: the rural network and land records.

• What is right about this objection

• Nothing measurable has come out yet. Hosting a meeting is not the same as getting results for farmers.

• Why it is still worth the effort

• Interoperability means different machines and systems can work together and share data. The rules for it are set early, in bodies like ITU's standards arm (TSB) [1].
• A country that is absent when the rules are written has to follow rules written by others later. That can tie its farmers to costly equipment from a few firms.

• India already has a working model of a federated, consent-based farm data system [5]. These Working Groups are where it can offer that model to other countries.

• How to judge it next year

• Did the Working Groups publish standards or guidelines?
• Do those texts reflect smallholder needs, such as cheap devices, local languages and farmer data rights?

12. What India Should Carry Into the Working Groups

• TEC should propose India's federated data model as a reference design
• AgriStack keeps data with the states and shares it only with the farmer's consent [5].

• Offering this as a model in the Working Group Terms of Reference [1] would make farmer data rights part of the global standard from the start.

• The Ministry of Agriculture should measure who the Farmer ID misses

• The Digital Agriculture Mission has an outlay of ₹2,817 crore, including a central share of ₹1,940 crore [6].

• Part of its review should count tenant and landless cultivators left outside AgriStack, since Farmer IDs are built from land records [5].

• DoT should link farm AI pilots to rural network coverage

• Hackathon apps [1] should be tested where only weak 2G/4G or no network is available, because rural access lags far behind urban access [7].

• ICAR should judge AI advisories by what changes on the farm

• Advice from tools like Bharat-VISTAAR [3] should be checked against real results: lower input costs, less crop loss, and how many small farmers use it again.

13. Anchors for Answers

• Data: In developing countries in 2020, 65% of urban people had internet access but only 28% of rural people [7]
• Data: More than 10.31 crore Farmer IDs were created under AgriStack by 3 Aug 2026, each linked to digitised land records [5]
• Data: The Digital Agriculture Mission has an outlay of ₹2,817 crore (central share ₹1,940 crore) [6]
• Law/Case: Digital Personal Data Protection Act, 2023 and DPDP Rules, 2025. AgriStack is built to follow them, with consent-based sharing of farmer data [5]
• Comparison: Sub-Saharan Africa: the World Bank finds that AI farming tools there are held back by weak connectivity, unreliable power and smallholders' inability to afford devices and data [8]
• Scheme: Digital Agriculture Mission / AgriStack, the domestic DPI (shared public digital systems) that any AI farm advisory, including Bharat-VISTAAR, relies on [3][5]

14. Mains Relevance

• GS-II: Important international institutions, their agencies and mandates (ITU, FAO, IFAD, WFP); India's role in multilateral forums.
• GS-III: E-technology in aid of farmers; science and technology developments; food security; reducing post-harvest losses and improving supply chains.
• GS-IV (minor): Ethics of AI, including fairness and inclusion of smallholders.
• Possible question stems: 1. "AI and IoT can transform Indian agriculture only if they are farmer-centric and interoperable." Discuss, with reference to recent global and national initiatives. 2. Examine the role of international standard-setting bodies such as ITU in governing AI for food security. How can India use such platforms for Global South leadership? 3. "Digital technologies can reduce post-harvest losses and improve traceability across agri-supply chains." Critically analyse the opportunities and constraints in the Indian context.

15. Related Topics to Study Next

• ITU structure (ITU-T/TSB, ITU-R, ITU-D): GI-AI4FS is coordinated through ITU's standardisation arm.
• FAO, IFAD and WFP (the Rome-based agencies): these are the initiative's partner organisations.
• India-AI Impact Summit 2026: this is India's main AI-diplomacy platform, and it had an agri-food session.
• Bharat-VISTAAR, AgriStack, Digital Agriculture Mission: these are the domestic AI and digital public infrastructure for farmers.
• AI for Good Global Summit and WSIS Forum: these are the UN venues where agrifood AI is debated.
• India Mobile Congress: the venue, and a sign of telecom–agriculture convergence.
• Post-harvest losses and food supply chains: a key area where AI could make a difference.
• IndiaAI Mission / AI governance guidelines: India's national framework for responsible AI.

16. NCERT Concepts to Revise

• Digital public infrastructure (Polity): AgriStack and Bharat-VISTAAR are shared public digital systems; the note treats them as India's farm-AI foundation.
• Digital Agriculture Mission (Geography): This mission builds the Farmer ID registry and land-linked data that India's farm AI services rely on.
• Digital Personal Data Protection Act 2023 (Polity): The note says AgriStack's consent-based sharing of farmer data rests on this law.
• Post-harvest losses (Economy): Cutting post-harvest losses and improving traceability is a named benefit of AI and IoT, and a likely Mains theme.
• Food security (Economy): The initiative's wider goal is food security, which is why GS-III questions on it are likely.

17. Common Errors / Trap Areas

• Wrong coordinator: the coordinator is ITU, not FAO. FAO is a close partner, and FAO's own release says FAO and ITU "established" it. [1][2]
• Wrong Indian host: the host was TEC under DoT, not MeitY or the Ministry of Agriculture. ICAR was the co-host. [1]
• First vs second meeting: the first meeting was at FAO Headquarters (June 2026), and New Delhi hosted the second. [1][2]
• Hackathon names: the PIB release names it both "AI for Agri-Food Systems" and "AI Innovation for Resilient Agri-Food Systems". Both refer to the same 8–9 Oct event. [1]
• IFAD vs IFPRI: the partner is IFAD, the UN agricultural development fund. IFPRI is a CGIAR research centre and is not involved. [1]