She is known as the AI Lady of India and for good reasons too. Last week, Ms. Anna Roy, Senior Advisor NITI Aayog, spoke on the AI Ethics framework, a discussion paper that the policy think-tank has created. An evolving framework is what will get us there – Responsible AI for all. She was addressing the Xperience AI Virtual Summit. 

As new frontier technologies become immersive and all-pervasive, there’s a confluence of sectors as well. Given the massive socio-economic consequences and deep inter-linkages, it necessitates that AI ethics in India is globally aligned. These technologies comprise an entire suite and the distinguishing line between the cyber and the physical world is blurring. Privacy infringement and its consequences are severe, and that’s why some of the large IT companies may not be pushing the boundaries of facial recognition technology to the extent they are capable of. The unintended consequences, at times, have been disastrous. As it becomes stark, law enforcement has a huge task cut out and needs to tread the path with utmost care.   

The National Strategy for AI was unveiled about two years back, and it charts out how Artificial Intelligence can enable higher GDP growth for the country. A recent NASSCOM study pegs the growth figure at 500 billion dollars. The idea of ethics in AI is something we all understand, but it can’t meander to a generalized approach, and that’s why this framework has been created. 

AI can impact the users directly and the population at large, indirectly. The direct impact is due to specific AI systems and leads to systems consideration. The indirect impact has societal considerations (such as) due to widespread automation, deep fakes, and the future of jobs that is going to change radically.

There are six kinds of systems consideration:

• Safe & reliable deployment 
• Explainability – why a specific decision was made
• Consistency
• Incorrect decisions can lead to the exclusion of certain sectors
• Accountability
• Privacy risks 

These considerations, unless addressed, can cause a variety of crippling challenges and lead us to the realm of irresponsibility. Also, they are nuanced and sector-specific. For example, if the state is going to take an administrative decision based on AI, then explainability is crucial. The decision will impact a million lives or more. There are legal provisions that already exist for some of these considerations, but given how dynamic these technologies are, policy-making is often playing catch-up. And, of course, we don’t want a very constricting environment that stifles the spirit of innovation. That’s why it’s like walking a tightrope. 

In India, we require specific guidelines, and the overarching principles can be used in setting them. The Draft PDP Bill which has to be passed yet, in the Parliament, is sector agnostic. These laws have to be put in place at the earliest. This is just one of the many examples used here to highlight the urgency.

From a societal consideration angle, the impact on jobs is something that’s been discussed, and both the pros and cons have been widely circulated. It’s up to us whether we want to focus on the jobs lost, or as WEF had estimated (pre-pandemic) the 133 million new ones that are going to be created by 2022. Okay, push it now by a year or two, but nothing changes otherwise. Are we doing enough to reskill and align ourselves to the future? Inasmuch, AI can also be put to malicious use such as psychological profiling on targetted groups to further one’s propaganda. The Cambridge Analytica scandal is all too common. The entire society can be severely impacted if we don’t create the guardrails now. 

Seven Principles are in motion – safety & reliability, equality, inclusivity, privacy & security, transparency, accountability, reinforcement of positive human values. It’s these guiding principles that will take us there, and they aren’t ad hoc. Much thought has gone into the framework, and considerations such as AI case studies globally, constitutional provisions, and even the concept of morality has played an important role. Again, the significance of global alignment cannot be over-emphasized. 

To end, it’s a dynamic field, and it’s imperative to invest in research and development, and though it sounds cliched, stakeholders need to come together. They are governments, citizens, regulators, the private sector, researchers & standard-setting bodies.   

Let us never forget the Peter Parker Principle – With Great Power Comes Great Responsibility.