Agriculture can be referred to as one of the oldest and primary occupations in human history and is the primary source of livelihood for about 58% of India’s population. Even though most people in India depend upon agriculture sector, its contribution to the GDP of the country is not significant. In fact, its year-by-year contribution in the GDP of India since independence has gradually decreased. This could be attributed to various challenges faced by the sector such as dependency on monsoon for irrigation, lack of knowledge in relation to time of sowing and quality of seeds, soil erosion, outdated farming practices etc.
Unlike other sectors, agriculture has been left behind when it comes to making use of the latest technologies. In this article, we will discuss the emergence of data-driven startups who are using artificial intelligence (AI) to script a change in this sector. These startups use AI to integrate technology with agriculture to provide the much-needed push to the sector. Since these startups rely mostly on data, they need to be mindful of the relevant data regulations applicable to them.
Data-driven AI startups
The current market in India is flooded with agritech startups trying to replace intuition and traditional methods of farming with technology. In the financial year 2019-20, Indian agri-food tech startups raised more than USD 1 Billion through 133 deals. Achieving any kind of automation using robots is still very expensive for a farmer in India. This is where AI comes into the picture. The resources required by a farmer to receive information collected and processed by these agritech startups, using machine learning tools, is minimal. It is very cost effective when compared to purchasing expensive hardware which would need to be updated regularly.
These startups are essentially trying to transfer the instincts and intuition of a good farmer to machines using AI. They collect and feed non-personal data sets taken from various sources such as on-farm sensors, IOT sensors, social networks, data from government organizations like ISRO, IMD etc., to the AI under development. The software can then be used for various use cases such as crop and soil monitoring, predicting agricultural analytics, building supply chain efficiencies, etc.
Regulatory regime governing data collection
As discussed above, these startups depend upon non-personal data sets to train their machine learning projects. In simple language, non-personal data is any data which is not personal data or data that is without any personally identifiable information. Non-personal data includes data sets pertaining to environment, production processes, geospatial information or any personal data set which is anonymised in a way that no individual can be re-identified from the data set.
While the personal data i.e., data by which a specific individual could be identified, is regulated by Information Technology Act, 2000 and allied rules, usage of non-personal data in India remains unregulated. The Central Government is deliberating on a personal data law which is currently in the form of the Personal Data Protection Bill, 2019 (Bill). The Bill also contains some passing references to provisions for regulating non-personal data. However, in what could be considered a positive step, the Ministry of Electronics & Information Technology constituted a Committee of Experts (Committee) on September 13, 2019 to deliberate on a data governance framework for non-personal data. The Committee has proposed deletion of the non-personal data related provisions from the Bill to harmonize the two prospective legislations. The Committee was formed with certain goals in mind: (i) to study various issues relating to non-personal data; and (ii) make specific suggestions for consideration of the Central Government on regulation of non-personal data. The Committee submitted its report on July 12, 2020 which sought feedback from public till August 13, 2020. After considering the public’s comments, the Committee released a revised committee report on December 16, 2020 (Revised Committee Report). The Revised Committee Report was once again opened for public comments till January 31, 2021. The Committee is yet to release an updated report post this.
The Revised Committee Report identifies that abundant availability of data is the primary driver of AI. It further categorizes non-personal data into two broad heads: (i) based on type – public, and private; and (ii) based on sensitivity. It also suggests setting up a separate non-personal data authority and explains key non-personal data roles such as a community acting as the data principal, data custodian, data processor, high-value data sets and data trustees etc.
All things considered, the Revised Committee Report discourages accumulation of data in the hands of few private players or government bodies and encourages data sharing for economic purposes. It understands that it is very important for Indian businesses to have open access to non-personal data sets collected by other private players or government bodies. The regulations pertaining to usage of non-personal data sets are yet to be framed. However, the Revised Committee Report gives a fair idea on how the regulations could positively develop in favour of small Indian businesses over time.
According to MarketsAndMarkets, AI in global agriculture market is expected to reach USD 4 Billion by 2026, growing at a compound annual growth rate (CAGR) of 25.5%, from USD 1 Billion in 2020. In India, where the market is still in its nascent stage, there are various startups that have come up like CropIn, Agnext, Zentronlabs, Intello Labs, Wolkus Technology Solutions etc. They have also been successful in attracting investments from various venture capital and private equity firms like Omnivore Partners, Nexus Venture Partners, Zoho, Kalaari Capital, Chiratae Ventures etc.
As discussed earlier, these startups rely heavily on usage of available data sets to improve their machine learning software. As on date, there is no legal restriction on these startups from using the available non-personal data sets to their benefit. The Revised Committee Report also gives a positive indication in relation to sharing non-personal data sets with businesses for developing their products.
Considering the market conditions and the favourable view adopted by the Committee, one could certainly hope to see agritech startups growing at an exponential rate. The untapped potential for growth in the agricultural space in India, should be another reason for investors to be bullish about agritech startups.
Ajay Joseph | Partner, Veyrah Law; Anshul Pandey | Associate, Veyrah Law