Hype and hysteria over artificial intelligence must be tempered to prevent knee-jerk reactions from regulators and to ensure India can reap the benefits cutting edge technology offers, write Abhivardhan and Mridutpal Bhattacharyya

To its boosters, artificial intelligence (AI) is the silver bullet to cure humanity’s many problems. Critics see the technology as an existential threat to our species. Amid the hype and hysteria, governments and global business are racing to make the next breakthrough, define the challenges AI presents and draw up rules of engagement for a technology that the UN University Centre for Policy Research says is “transforming the geopolitical order” with a “shift in the balance of power between intelligent machines and humans” already visible.

The heated world debate over AI’s potential – both good and bad – risks sparking not only an arms race for strategic leadership, but also a patchwork of unwarranted and stifling regulation and a public loss of trust in the technology. And all of this at a time when the world needs more, not less, technological innovation.

AbhivardhaN, Indic Pacific, Deconstructing AI
Managing Partner
Indic Pacific Legal Research

Recognising both the hyperbole over AI’s potential in the near future as well as the hysteria driving public fear of the technology is a vital first step to policies that can control the dangers while allowing for its continued evolution.

Governments in the global south, unlike many developed economies, have yet to formulate concrete policy and regulatory approaches to AI. Unlike many of its peers, however, India’s massive and thriving tech sector is already heavily involved in developing AI – including through ventures with global technology companies. This makes it vital that India filter out the noise and focus on shaping national priorities that will benefit regulators and courts, as well as innovators and entrepreneurs.

Global sectoral implications

The hyperbole can be traced to two main sources: marketing of AI technology by self-interested investors (also known as the “Google effect”, after CEO Sundar Pichai’s 2018 declaration that AI is “probably the most important thing humanity has ever worked on”); and its promotion by tech moguls as the answer to the problems plaguing humanity.

The hysteria, too, has two main roots: apocalyptic warnings that AI poses an existential threat to humanity; and its potential to create mass unemployment through automation of jobs.

One 2013 estimate put potential job losses in the US at up to 47% of the workforce over the coming decade or two. But in 2019-2020, rather than mass unemployment, jobless rates in advanced economies were at historic lows. Ironically, the evidence suggests that between 1999 and 2010, automation led to a net 1.5 million new jobs in Europe alone.

At the same time, labour productivity has continued to decline – with the UK’s 10-year average labour productivity growth since 2007 the lowest since 1761. Even leading conglomerates that stand to benefit the most from AI have not seen a revolution in productivity.

So why haven’t the supposed threats from AI materialised? One reason is that adoption has been slower and more expensive than expected. And one outcome of that has been a boom in pseudo-AI.

As The Guardian newspaper reported in 2018: “It’s hard to build a service powered by artificial intelligence. So hard, in fact, that some startups have worked out it’s cheaper and easier to get humans to behave like robots than it is to get machines to behave like humans.”

Second, innovation in AI is getting more difficult and is mostly aimed at fine-tuning existing products rather than introducing radically new ones. Early gains from machine learning have already been harvested, and diminishing returns appear to have set in.

Third, it is not especially profitable for businesses to actively invest in AI, considering the slow growth in demand and limited applications in most Western states. The biggest innovation we have seen is in visual systems for autonomous vehicles, which are particularly notable for not being on any roads – which looks likely to be the case for some time to come.

India has set out high hopes for AI, forecasting the technology would add about 15% to the gross value of the economy by 2035. Energy, retail, manufacturing, healthcare, education and agriculture were among sectors seen benefiting, with the creation of numerous jobs.

As part of the government’s “Digital India” drive, Software Technology Parks of India (an organisation established by the Ministry of Electronics and Information Technology) aims to create centres of entrepreneurship nationwide to promote AI research and development.

The DRDO Young Scientist Labs – established in 2020 by the government at Bengaluru, Mumbai, Chennai, Kolkata and Hyderabad – seek to make India self-reliant in AI.

Many big companies are already engaged in AI innovation, tying up with universities and government bodies. In September 2021, Wipro GE Healthcare partnered with the Indian Institute of Science (IISc) to build a lab to work on healthcare diagnostics, deep-learning technology, machine learning and AI systems.

In 2020, Accenture opened up an AIinnovation hub in Hyderabad with more than 2,000 personnel to innovate solutions for security, automation and blockchain domains.

In partnership with IIT Hyderabad, Nvidia also established a lab to accelerate research and innovation in the commercial adoption of AI.

The IBM Centre of Excellence was set up in collaboration with the government e-marketplace (GeM) in 2020 to infuse AI into software and systems. IBM has also partnered with the Karnataka government and the NITI Aayog to focus on agriculture and the environment.

These and other developments show how India’s thriving ecosystem and large talent pool make the country an attractive destination for setting up AI labs and research centres.

However, while India is developing AI for application in a wide range of sectors, the legislative framework has not kept pace. The country has yet to see any legislative framework on AI or algorithms, or even, for that matter, a multi-sector AI regulation framework that addresses the many types of the technology that can be applied in different industries.

Regulating disruption is a challenge for the government, since sector-specific or industry-specific needs and requirements can drastically shape the role of AI technologies. How those needs emerge is not subject to regulation. However, the government can surely try to keep up to speed on the issues, and can contribute to the development of disruptive technologies by investing wisely in R&D.

For instance, India allocated almost 25% of the 2022 defence budget to R&D. While this was a welcome move, the government must also invest more in other areas too. It is through experience of technologies over time that policy and regulation will be smooth, reasonable and strategic.

Examining repercussions

In its approach to regulating technology, the European Commission looked not just at the hype of AI, but also the repercussions of specific technology-related disruptions created by multinational companies (especially tech giants). Yet even the commission’s competition law approach against companies like Amazon has not fared well.

The Competition Commission of India has also taken a competition law trajectory for tech companies. However, the approach is not mature enough because policymakers and authorities need a better understanding of the intrinsic potential of disruptive technologies, and how sector-specific developments in such technologies take place.

There is no doubt that as companies in India put AI to use, their business practices will at times become unacceptable or unsustainable. This raises the risk that India will take a too-rigid approach to AI. The government must therefore ensure that as it tests and develops its regulatory approaches, it tackles business and technology practices from a similar plane, for example, with regulatory sandboxes.

The coming years will see a dire need for effective dispute resolution mechanisms, along with better judicial pronouncements by the Supreme Court and high courts, to shape AI regulation. For this to happen, it is vital that the hype surrounding AI is brought under control and that the sector-specific priorities of AI drive the evolving regulatory landscape.

Abhivardhan is the managing partner of Indic Pacific Legal Research and Mridutpal Bhattacharyya is the deputy chief innovation officer of the Indian Society of Artificial Intelligence and Law, a technology law research organisation specialising in AI ethics and legal regulation.

Law.asia subscripton ad red 2022