India wants to be on the cusp of artificial intelligence but lacks the laws to back it up

• Research in artificial intelligence (AI) has grown seven-fold since 1996 but India still lags woefully behind, according to a new report by Tata Consultancy Services (TCS).
• The Indian government does not have regulations to govern the development of AI and existing laws are ambiguous.
• India's private sector — Wipro, Infosys, TCS, and Tech Mahindra — also lag behind their global counterparts.

Most people use artificial intelligence (AI) at least once a day to help them carry out a task. Even something as basic as Google Search has an AI-powered algorithm behind it to make it more efficient.

Machines are only set to get smarter as more applications of AI come to light. Research around AI has grown by seven-fold since 1996, according to a study by Tata Consultancy Services (TCS). Yet, India lags far behind.

“We have one of the best technology talent pools in the world. If we fast-track and balance our progress on innovation, IP management, and entrepreneurship, we can realize the potential to become a global AI powerhouse,” said Santosh Mohanty, the Global Head for Components Engineering Group at TCS.

Lacking in research

Most of the research in AI over the past couple of years have been from scientists in the US, Europe, and China. In fact, the Chinese Academy of Sciences (CAS) holds the largest patent portfolio that deals explicitly with deep learning — a subset of AI where the focus is on recognizing patterns by mimicking how neurons interact in the human brain.

It's hardly surprising since out of 22,000 PhD researchers around the world, only 386 are from India, according to the Global AI Talent Report 2018.

Lacking in laws

Adding fuel to the fire, India also has no guidelines to govern the development of AI.

Until 2002, computer-related inventions were deemed ineligible for patents. Even though that has now changed, the existing laws pose their own challenges — like being too ambiguous and vague.

For instance, an algorithm can't be patented until it has a practical application or use case — even if it's solving a problem behind the scenes.

Lacking in innovation

Even though Infosys, TCS, Wipro, and Tech Mahindra are tech giants in the Indian software industry — they too lag behind their global counterparts when it comes to AI. Currently, the top five patent applicants are IBM, Microsoft, Toshiba, Samsung, and NEC.

All of them, except for IBM, are focused on computer vision — a form of machine perception that used deep learning to identify objects, videos, and images. IBM, on the other hand, is focused on natural language processing — like chatbots.

Alphabet, Google's parent company, is second only to the Chinese tech giant — Baidu — in owning portfolios of patents related to deep learning.

If India wants to create a $1 trillion digital economy, it has a long way to go in enabling its software engineers and scientists to be at the forefront of AI.

See also:
India is all set to deploy facial recognition but there is no law in place to keep a check

Here's what global tech CEOs have to say about India's data protection laws

Data localisation won't help 'local' firms or improve privacy — but it could cost India $43 billion