Powering India’s AI push through sensible data localisation

Representative Image

The 1939 Phony War was theoretically about Poland but was really about world domination.  The debate on data localisation is similarly phony.  Supposedly about privacy, individual freedom, and narrowly defined national security, it is actually about money and power. It is about a choice India faces -- lead the emerging digital world, or be a receptacle for "trickle down" technology.

Sensible Data Localization can help ensure India’s economic future by nurturing enterprises that can be global leaders in the "new" world economy driven by Artificial Intelligence.  A strong economy almost naturally secures India’s position in the world order.

First, the red herrings of privacy and national security.  "Developed" markets limiting cross-border data is protecting privacy, but India doing so becomes an assault on privacy.  Oddly, the loudest voices against India’s data localization initiatives are the ones that have provided data and helped companies like Cambridge Analytica cast an undue influence on our politics. The simple solution is to frame rules to govern data security and privacy, and an independent authority to ensure compliance.

The second phony issue is national security, specifically, the availability of data needed by our security services.  The solution again is simple – specified data, as well as data requested in line with established laws, must be made available to law enforcement agencies.  Onshore or offshore, data owners will have to meet specified delivery timelines or be shut out of the Indian market.

Which brings us to the real issue -- money -- who earns it, whose shares rise, which country sees new entrepreneurs develop and the direct result -- international power.

So why does data and its storage location become so important? Artificial Intelligence.

Still new, AI is revolutionising the world.  It will change the way we live, produce, think and govern ourselves. Exaggerating (maybe?), Putin said whoever leads in AI will rule the world.  AI, and the related concept of machine learning are based on the ability of the "machine" to predict.  Conventionally, predictions --the behaviour of cricket pitches, corporate actions or Kim Jong’s next missile, are made based on data of past behaviour, and linkages between those behaviours.

Substitute the human mind with its limitations on information storage and processing, with a machine -- ever hungry to "process" and store more info.  This machine -- AI, feeds, grows and evolves on one food -- DATA.  Data -- its quality, quantity and access drive success in the age of AI.  In fact, beyond a point the machines (as we have seen with computers playing chess) will seek and incorporate new data by itself -- constantly improving.

India, with its population, massively used infrastructure, mobiles, e-commerce and now smart cities has humongous DATA volumes.  The developed world has capital and an ecosystem to support building core competence in AI–research, education and large companies to test emerging technologies.

How do we use our Data wealth to bring their capital into India and build up our own AI research-entrepreneurial ecosystem?  Sensible data localisation, as part of a comprehensive digital policy.

• Harmonised, simple and clear digital policy regime – India’s digital policies are fragmented. We need a simple, clear and harmonized policy regime where the premium is on encouraging digitisation and building an ecosystem to make India a technology leader, not controlling and monitoring information flows.
• Choose the right data –Strategically choose data to localise. Medical, financial, e-comm, and government-services related data like traffic; benefits programs; and census data should be localised. Criteria should be privacy protection, and utility in fostering AI.
• Ensure access to data – Data must be universally available to create new applications. Infact, entrepreneurs should get subsidised access to data.
• Localisation is not non-transportability - Data can be bought for offshore use. However, this should be priced/taxed to "incentivize" players to take up application development in India.
• Protect economic value of data from government /PSUs – Data from government-linked sources should be treated as a national resource. A National Data Fund can be formed to monetise government-linked data.  The fund will "price and sell" data. After retaining part of the proceeds, the remainder will be paid to data generating entities. Retained proceeds can be used to make investments in AI supporting infrastructure, and potentially provide seed capital for AI ventures.

Early this century, China "forced" local startups in social media and e-commerce by banning global players. The objective was to save the Chinese market for local players – and what a market it was.  Today Baidu, Alibaba and Tencent have a combined value of $1 trillion+.  Data localisation will have a similar virtuous impact in building an AI ecosystem in India.

AI needs Indian data to grow.  With localisation, there will be investment inflows -- capital and technology, to tap this data.  The capital and access to technology will help startups come up.  As startups succeed, there will be more investment. India will gradually become a global AI hub.  In addition to the economic benefits, the technological edge will increase India’s influence on the global stage.

Centuries ago, "global" traders came to India seeking raw materials. The resulting colonial pattern of trade supported European industrialisation.  India was left behind.  The AI revolution will expand exponentially faster than industrialisation did, and create unsurmountable inequalities between nations. Sensible data localisation will ensure we are not left behind in this revolution.