Libertatem Magazine

Understanding AI-Enabled COVID Testing in India

Contents of this Page

This article focuses on the significance of Artificial Intelligence during the Novel Coronavirus (COVID-19) pandemic. This article will shed light on the government’s stance on AI and its usage to battle COVID-19.

How is COVID-19 Currently Tested in India?

India has one of the lowest ratios of testing in the world. It statistics stand at approximately 18 people per million. The nation is heralded the ‘Pharmacy of the World’. However, it falls short in the production of medical equipment. Specifically, it lacks adequate testing equipment. The absence of regular clinical check-up culture in India partly contributes to it.

India is a developing nation. Hence, it is unable to conduct a huge number of tests like its Western counterparts. The nation cannot bear the cost of these tests, especially when it already slipping into an economic downturn. This is where AI-enabled testing shines.

Polymerase chain reaction (PCR) tests are the most commonly used tests to detect the disease. But PCR tests are labour intensive and create up to 30% false negatives. In addition to this, symptomatic individuals who have tested negative will have to wait 2-3 days before getting a second PCR test.

Individuals showing symptoms with a negative PCR can be screened via radiology instead. It results in quicker diagnosis and considerable cost reduction. Further, it will reduce the chances of community spread by the way of false negatives.

Radiology tests can be a practical alternative to undergoing a PCR test for a second time. This can assist in deciding whether to quarantine the individual in the hospital/institution or send them to home quarantine. This will reduce the number of PCR tests administered. Additionally, it will open up resources in hospitals for those who need the same.

How Can AI-Enabled Tests Help?

This is where start-ups like 5C Network gains importance. 5C makes radiology more accessible. It allows hospitals and diagnostic centres to upload scans to the cloud. 5C’s AI algorithm excerpts the data and analyses the scans in the cloud. It then relays back the required information. 

Unfortunately, India has a very low ratio of radiologists – 1 for every 100,000. But, this situation makes the proposed solution suitable. AI radiology tests prove to be fast and scalable. It is because 5C Network offers Remote Radiology support and real-time analysis of reports by multiple experts of the Radiology Response Team. 

Researchers at King’s College, Massachusetts General Hospital and health-tech company ZOE have developed an AI diagnostic system. This system can calculate whether someone is probably COVID-19 positive based on their symptoms. The AI uses data from the COVID Symptom Study app to forecast infection. It compares the user’s symptoms with the results from conventional COVID-19 tests.

This method will aid impoverished populations with limited access to testing. Moreover, individuals who get a warning from the app may be tested conventionally thereafter. The government can incorporate such an AI model with the Aarogya Setu app. This would ensure success due to the sheer size of the Indian population’s dataset.

What is India’s Legal Viewpoint on AI?

India currently has no law that governs the realm of Artificial Intelligence (AI). Thus, the nation has to make do with existing laws and guidelines to regulate Artificial Intelligence systems. For the same, we will divide AI models into two:-

  1. AI models with human-assisted input
  2. AI models without human-assisted input

For the first model, the AI uses complicated matrices to analyse the datasets. It works on a programme created by human programmers. The second model also works on a source code that is written by human programmers. But, the difference is that the second model is able to deviate from its source code. At times, it makes automatic improvements to its own code without human interference.

Thus, many obstacles arise while trying to pin liability on AI. Suppose an AI-enabled test gives out false results that curtail treatment to an individual. In such a case, who can the individual recover the costs from? Will the liability fall on the AI developer, the intermediary using it, or the health staff authorised to use it?

NITI Aayog published ‘National Strategy for Artificial Intelligence’ in June 2018. It is the primary document that lays down what the Indian government’s stance on AI is. However, it has no binding effect besides the fact that it is a very comprehensive document on AI.

What is the Liability of These AI Models?

When it comes to the first model it can be easy to transfer the liability to the developer. But one must also consider the fact that the intermediary and the doctor using the systems were also aware of the risk involved. If the underlying risk was never conveyed, redressal may be sought through Section 18 of The Indian Contract Act, 1872 or under the provisions of the Consumer Protection Act, 1986.

The second model complicates the matter further. It creates a single solution to solve such a perplexing situation, which is to enter into contracts with health-tech companies that specify the liability. Under this model, redressal will flow through the civil court system via provisions of the Contract Act.

Does India have its own Data Privacy Laws?

We live in an era where ‘Data is the new Gold’. But, India has not framed any data protection laws yet. The Data Protection Bill, 2018 has only been laid down before the Parliament. However, the Bill appears to be a toothless tiger when it comes to dealing with artificial intelligence.

That being said, we must now try to regulate data privacy with the help of existing statutes. The primary piece of legislation that regulates the cyberspace is the Information Technology Act, 2000. Section 43(a) and Section 72 of the IT Act lay down a comprehensive framework for the protection of personal information. Section 72 penalizes the breach of one’s confidentiality and privacy. The offender may be punished with imprisonment up to two years, or with a maximum fine of ₹1 lac, or both.

In addition to this, the sensitive personal information rules issued by the Ministry of Communications and Information Technology penalise unauthorised sharing of sensitive personal information. In this case, sensitive personal information encompasses one’s medical information.

Furthermore, under the Indian Medical Council (Professional Conduct, Etiquette and Ethics) Regulations, 2002, health officials are obligated to safeguard the privacy of patients during all the phases of the treatment process. This includes all information communicated by the patient to the doctor. This information may also concern their personal and domestic lives.

How does the Future of AI seem in India?

Introducing AI in an already chaotic climate makes everything more complicated. Under Indian jurisprudence, AI does not constitute a separate legal entity. How can the judiciary punish an AI when it shares sensitive personal information of an individual? Interpreting the IT Act drives us to the conclusion that the developer or the facilitating organization must be at fault. However, Medical Ethics puts the onus on physicians to ensure that the personal information of patients remains confidential. It is important to note that the Medical Code lacks any robust enforcement mechanism.

A reasonable conclusion herein would be to hold all the contracting parties accountable. Yet, the same raises new questions. For instance, to what extent must each party be held liable? How will the extent of this liability be decided upon? The responsibility of answering these questions falls on our government  – after all, if not to frame adequate laws, then what is the primary goal of the legislature?

Author’s Suggestions

The suggestions below reflect the personal views of the author. They aim to fill this lacuna in law:

  1. Enactment of an AI-specific Act that lays down the regulatory framework of AI in India.
  2. The inclusion of AI-related aspects in the Data Protection Bill, 2018. The same must address many of the privacy-related queries raised in the article.
  3. Mandating arbitration clauses in AI-related contracts. The same will shift the burden from courts until we have a robust AI regulation. Moreover, lawyers experienced in the field of AI would be in a better position to arrive at judicious conclusions.
  4. For the time being, AI-enabled testing for COVID-19 must be only used in combination with conventional tests. The same will redirect much of the legal problems that may have otherwise arisen.
  5. The recently unveiled National AI portal must host chatbots to address queries of the public. The government must expand the platform to register AI-enabled systems, forming an AI registry.
  6. Mandate Indian developers to incorporate The Global Initiative on Ethics of Autonomous and Intelligent Systems of the IEEE’s chapter on ‘Personal Data and Individual Access Control in Ethically Aligned Design’. 

Conclusion

It is inevitable that Artificial Intelligence plays a pivotal role in the future as well as in combating COVID-19. For now, the government must harness the power of AI to help in the testing of the Novel Coronavirus and in rebuilding the nation.

For the future, India should take note of the best practices that evolved at the global stage pertaining to electronic health and medical records maintained by AI systems. Then, it should apply the same to the Indian populous by tailoring it to Indian needs. The power of AI will definitely help boost our economy in a post-Coronavirus world. 


Libertatem.in is now on Telegram. Follow us for regular legal updates and judgements from the court. Follow us on Google News, InstagramLinkedInFacebook & Twitter. You can also subscribe to our Weekly Email Updates. You can also contribute stories like this and help us spread awareness for a better society. Submit Your Post Now.

About the Author