Several years ago, I watched a movie titled Her. It tells the story of a writer who uses an artificial intelligence (AI) system to help him with his work. Eventually, he falls in love with it and goes on to have his heart broken. In 2014, the premise of this movie seemed far-fetched. In retrospect, I realise that it actually displayed immense foresight.
Take, for instance, Anima — a virtual AI companion with whom you can have a “friendly chat, roleplay, grow your communication and relationship skills”. Anima claims to help its users by modelling a virtual friend, one that you can rely on in tough times for a quick chat. So intimate and real are these virtual friends/partners that a woman in New York married her AI-generated husband earlier this month, much like in Her. This is only one example of the vast human-to-human societal changes that AI is posing. Technologies such as Anima, which will only become more “real”, have a wide-ranging impact on laws pertaining to data privacy, security and evidently, even marriage.
The rapid pace of technological developments being thrown at governments and citizens is staggering. India too, is in the midst of its own digital revolution. Technology enthusiasts aside, even lay people are relying on generative AI platforms to plan holiday itineraries and meet essay deadlines. India’s open-arms response to ChatGPT has been appreciated by the OpenAI founder, Sam Altman, who was recently in the country and met Prime Minister Narendra Modi. Even independently, the Indian government is actively pushing for a digital transformation with its ‘Digital India’ initiative. It has introduced the Digital Personal Data Protection Bill, 2022. The Digital India Act (DIA), being called a landmark legislation, is also in its final stages. According to a report in The Indian Express, one of the key provisions under the Bill would be that all online fact-checking platforms must register with the Centre. Whether you love these pieces of legislation or hate them, it is undeniable that they mark the giant strides that the country is taking in the digital space.
The DIA is set to replace the rather outdated Information Technology Act. It is interesting to note the sheer vastness of what it proposes to cover. Especially the way it seeks to tackle the serious harms of emerging technologies such as deepfakes and AI-related misinformation. The Union Minister of State for Information Technology has announced that the DIA will be largely “principle-based”, a policy that will ask technology companies to regulate themselves based on broader principles that the law will establish. This is major signalling by the government, and it possibly cements its promise that technology regulation will not trump technology innovation.
It will be fascinating to observe how governments will structure technology-related regulations. Lawmakers find themselves at another perplexing crossroads, one that goes beyond the usual debate of ethics v advancement through AI. This is because technology law experts have delivered a split verdict with regard to AI regulation. Some say legislation is required, albeit of a particular kind. Others question if legislation on the technology is even fruitful at this stage.
The school of thought that proposes that it is too early to legislate argues AI technology is still at a nascent stage. Nonetheless, what is clear is that it is becoming increasingly unfeasible for governments to permit life-altering technologies to be introduced to the world without some systems in place to insulate users. Thus, the most effective path now would be for countries to regulate such technology, as is being done by the European Parliament. But, any regulation proposed or implemented must be broad enough for it to be constantly tweaked as technology progresses.
This raises an unusual constitutional conflict, as it challenges the very notion on which laws are customarily drafted — to ensure they are water-tight and exhaustive. However, if laws on this subject are not designed to be flexible, and accommodate new evolutions, they will become redundant even before they are effectively applied. However, even whilst pushing for openness, it will be key for governments to ensure that simpler and less-detailed legislation does not blur the lines that protect citizens and take unconstitutional form. Lawmakers will be required to exercise greater precaution, considering that regulations associated with emerging technologies are becoming increasingly linked to our essential freedoms.
With people being cognisant of the ramifications of AI on fundamental rights, efforts are being made to toe the line on the freedom given to AI-based systems. For instance, the Federal Trade Commission (FTC) in the US has issued a statement about the usage of automated systems that make critical decisions like hiring, awarding of loans or access to housing.
Imagine it: An automated system utilises existing datasets to make decisions regarding a job opening for accountants. Now consider that in the past only male candidates have been selected for this job. Evidently, such a dataset would be severely imbalanced, and result in skewed decision-making.
To understand the troughs of emerging technologies and then consequently provide for them in our laws, what continues to be of utmost importance is dialogue between experts from cyber security, engineering, technology, and legal communities. This promotes much-required transparency as it assists parties that utilise such software in recognising the fallacies and gaps that exist. This will, in turn, ensure that laws being introduced to regulate technology, such as the DIA, come with inbuilt safeguards, and importantly, with the flexibility to build them in.
The writer is a lawyer and incoming LL.M. candidate at New York University
