Law in the Time of Artificial Intelligence: Who is Alexa? by Anuradha R. Chowdhary

Introduction

In 2001, Steven Spielberg and the legendary filmmaker (and magic-creator) Stanley Kubrick brought ‘A.I. Artificial Intelligence’ to our screens. Originally conceptualized by Kubrick, the film is about an android, who appears to be a little boy, programmed with the ability to love. Leaving the audience questioning whether robots can love, the film also rebooted the age-old question proposed by Alan Turing (the great scientist) in the 1950’s i.e. ‘can machines think?’. Hollywood followed with several fear-inducing and thought-provoking films on the subject and ignited our minds to the idea of being governed by and/or being destroyed by robots.

With the arrival of the 4th industrial revolution, Turing’s question is relevant even today, more so now than in 2001. All set to alter the way we live and interact with each other (and with our devices), the acceleration of the 4th industrial revolution will unshackle our lives. Artificial intelligence (AI) is no longer a question of the past and we can’t help but wonder – if machines really can think then can they be held liable?

Explaining Artificial Intelligence

AI relates to technologies that are designed to mimic human actions and behaviour, look for patterns, learn from experiences, figure out how to respond to situations based on learnings and even predict outcomes. AI significantly impacts economic growth and transforms businesses by providing ease in conducting and protecting the same. Because of the ability of AI to perform human tasks without human involvement, the impact of AI on occupations will be exponential. The main contribution of AI (so far) has been in the finance, automotive and telecommunications sectors.

The advancement and transformation of AI is inevitable and the ability of the law to catch up with AI or rather to adapt to AI is still in process. Elon Musk’s new venture ‘Artificial General Intelligence’ is currently in development to ‘rival and surpass the cognitive capabilities of humans’ and to replicate the human brain[1]. However, the outcome of AI simulating any biased patterns can be extremely dangerous and biases may even fortify over time. Since human actions and behaviour form the very base of AI, the need for AI to be a personification of unbiased persons is paramount.

Hey Alexa

If you’re familiar with the Internet of Things (IoT) then you’ll be familiar with the disruption caused by technology in our daily lives. If you aren’t familiar with the IoT (which you should be), well then, IoT is basically when your device comes with a connectivity feature that makes the device smarter than it would be on its own.

For example – a watch that can connect to the internet is a smart watch. Amazon’s Alexa is a smart speaker device that acts as a virtual assistant. You can ask Alexa questions, check the weather, check your stocks, play podcasts, order your groceries and even control your other smart devices through Alexa. However, as much as Alexa hangs on to every word her owner says, she also records every word her owner say. The data collected by (intrusive) smart devices such as Alexa, undergoes ‘data mining’ which is a process by which AI discovers and analyses patterns in large data sets. The data then collected is used to tailor advertising and promotions to consumers, spot sales trends and to essentially invade your privacy. While this may seem like an unlikely danger to your day-today life, the security risk associated with the interference of devices and unauthorized surveillance is colossal and unresolved.

Robot Liability

Defects in robotic software can cause irrevocable accidents. Let me give you an example. A factory owner rents a recycled AI-based robot. Since AI is self-learning, the robot may have retained certain patterns it had previously acquired and accidentally apply the same to its current environment. While reprogramming is possible, one cannot ascertain the effectiveness of the same. In an accident caused by the recycled robot, will the liability lie with (a) the manufacturer for not properly reprogramming the robot; or (b) with the factory owner for being negligent and not properly monitoring the robot; or (c) with the robot itself.

Currently, the Sale of Goods Act, 1930, the Consumer Protection Act, 1986, the Indian Contract Act, 1872 and statutes specific to certain products (such as automobiles, food, etc.) regulate product liability in India. However, AI and robots cannot with certainty be classified as products and therefore the laws that regulate product liability cannot be specifically applied to AI or used to adequately resolve the challenges of AI based robots.

Personhood and AI

AI has arrived in India but currently robots do not enjoy the status of a legal ‘person’. The essence of personhood lies in the right to own property and the capacity to sue and be sued[2]. Globally, the legal personality of AI falls under a grey area and a uniform consensus is absent.

In the U.S., Uber’s self-driven vehicle killed a pedestrian in Arizona and the liability issue was settled (and deeply buried amidst the legal chaos that ensued in the aftermath). In Saudi Arabia a robot was given citizenship. Earlier this year, the Kerala police added a robot to their force and Chennai got its second robot themed restaurant. A restaurant in Bombay has robots serving shots of alcohol.

In 2018, a cardiologist in Ahmedabad performed the world’s first in-human telerobotic coronary intervention on a patient 30 kms away. While the surgery, which was technically performed by a robot and supervised by the cardiologist, was successful, in the high chance that it failed, who would the patient’s family hold responsible? Without having laws to hold AI responsible, it is difficult to ascertain who would take responsibility for any damaging actions. Further, in the absence of recognition of personhood of robots and AI, the interpretation of liability is unsettled.

Conclusion

AI, which is no longer limited to science-fiction, has become a reality and development of laws around the world are in process to regulate it. The U.S. has initiated laws to manage AI, Korea and Japan have issued clarifications on the protection of human life over products and the EU has adopted the General Data Protection Regulation (GDPR) to protect the personal data of its citizens.

India too is starting to notice the risks associated with unregulated AI. According to Dipak Misra (former Chief Justice of India), “the future of any new-age technology lies in the regulations that govern them”. NITI Aayog, which is the Government of India’s policy think tank, has recommended including AI in the scope of our intellectual property regulatory regime, formulating a national data privacy law that matches the standards of the GDPR and creating sector specific guidelines for data privacy and security. The 2019 Union Budget also announced plans to set up a national centre for AI and launch a national program on AI.

Unregulated AI is certain to pose a risk to the existence of mankind. Everything that bears any impact on human existence including modes of transport from bikes to aeroplanes, food, drugs, medicine, etc. have been regulated. However, effective methods to manage AI surveillance, clarification of questions pertaining to liability and the notion of robots enjoying the legal status of ‘person’ are still distant ideas, and while the scope of AI continues to grow, the need to regulate the outcome of its actions remains pending. The best way forward for India to ride the AI wave and thrive in this digital world is to (quickly) formulate and adopt innovative and enterprising regulations for AI.

“In a few years, that bot will move so fast you’ll need a strobe light to see it.  Sweet dreams…”

– Elon Musk.

Bibliography

The next industrial revolution is already here, and it’s in your home. (2019). Retrieved from https://www.cbsnews.com/news/internet-of-things-iot-next-industrial-revolution-davos-2019-world-economic-forum/

The Fourth Industrial Revolution: what it means and how to respond. (2019). Retrieved from https://www.weforum.org/agenda/2016/01/the-fourth-industrial-revolution-what-it-means-and-how-to-respond/

What Is The Real Difference Between Automation And AI?. (2019). Retrieved from https://becominghuman.ai/what-is-the-real-difference-between-automation-and-ai-366513e0c910

AI, automation, and the future of work: Ten things to solve for. (2019). Retrieved from https://www.mckinsey.com/featured-insights/future-of-work/ai-automation-and-the-future-of-work-ten-things-to-solve-for

Artificial Intelligence and Law – Law Times Journal. (2019). Retrieved 19 July 2019, from https://lawtimesjournal.in/artificial-intelligence-and-law/

Clifford, C. (2019). 9 of the most jaw-dropping things Elon Musk said about robots and AI in 2017. Retrieved 22 August 2019, from https://www.cnbc.com/2017/12/18/9-mind-blowing-things-elon-musk-said-about-robots-and-ai-in-2017.html

Elon Musk’s plan to replicate the human brain with AI just received $1bn from Microsoft. (2019). Retrieved 22 August 2019, from https://www.independent.co.uk/life-style/gadgets-and-tech/news/elon-musk-ai-openai-microsoft-artificial-intelligence-funding-a9016736.html

What is Product Liability? – FindLaw. (2019). Retrieved 29 August 2019, from https://injury.findlaw.com/product-liability/what-is-product-liability.html

India: Product Liability – AZB & Partners. (2019). Retrieved 29 August 2019, from https://www.azbpartners.com/bank/india-product-liability/

White, M. (2019). Big Data Knows What You’re Doing Right Now. Retrieved 30 August 2019, from https://business.time.com/2012/07/31/big-data-knows-what-youre-doing-right-now/

(2019). Retrieved 30 August 2019, from https://www.niti.gov.in/writereaddata/files/document_publication/NationalStrategy-for-AI-Discussion-Paper.pdf

Spielberg, S. (2001). A.I. Artificial Intelligence [Film]. Hollywood.

[1] (“Elon Musk’s plan to replicate the human brain with AI just received $1bn from Microsoft”, 2019)

[2] L. B. Solum. Legal Personhood for Artificial Intelligences. North Carolina Law Review, 70: 1231–1287 (1992)

The easiest Rubik’s Cube solution is available here in many languages. Learn it in an hour to impress your friends.

[/vc_column_text][/vc_column]

[/vc_row]