EU’s landmark AI Act has the world watching with a hawk eye, ET LegalWorld

The world’s first legislation designed to regulate artificial intelligence has been announced by the European Union on Artificial Intelligence (AI), a comprehensive law on AI made by any other jurisdiction. The law comes with hefty fines for developers who fail to comply.

A landmark legislation, the European Union’s AI Act has a reach that extends beyond the territorial remit of its 27 member states and contains predetermined punishments up to $37 million.This one of a kind legislation and will be closely watched by other jurisdictions to be adopted by them in the coming years. This AI Act has been revered by all countries who are monitoring its application of solid regulations on AI users, applicators and makers.

EU’s AI ACT and India

India is not currently in the process of making separate legislation for regulating Artificial Intelligence like Europe, Brazil, etc. However, the recent advisory by MeitY created an anticipation of the policy direction India would take as we move forward with specific regulations governing AI.

However, the lack of separate legislation does not mean that the existing regulations and upcoming digital laws will not apply to AI technologies. It is crucial to understand that these laws, while not specifically designed for AI, can still have a significant impact. Before we discuss the importance of India adopting separate AI regulation, it is essential to emphasise the need for these existing and upcoming laws to work in harmony, as they govern various facets and functions of AI technologies.

“For instance, as contoured during the Digital India Act (DIA) consultation, the government is considering having provisions within DIA that would define and regulate high-risk AI systems. Therefore, as these upcoming laws separately handle various concerns with AI solutions, more effort is needed to establish coordination between laws so that different building blocks work in tandem to tackle the harm posed by the technologies,” said Kazim Rizvi, founding director, The Dialogue.

Moreover, it is imperative for the government to step in and provide soft-touch prescriptive directions for the industry. “These directions should focus on operationalising some of the trustworthy AI principles, such as transparency and explainability, accountability, fairness and non-discrimination, Reliability and safety/security, human autonomy and determination, privacy and data protection, social and environmental sustainability, governance and oversight, and contestability,” said Rizvi.

Experts think that the governments could follow normative theories of regulation and institute market mechanisms such as a (a) audit of features for AI developers and AI deployers based on the principles mapped for them and (b) a market for principles-based accreditation, enabling a competitive edge for platforms. These actions are not just desirable, but necessary for the effective regulation of AI in India.

Artificial Intelligence Act (AI) is the next level the world is aiming to act to ensure innovations yet ensuring privacy and regulatory concerns. The European Union has come out with a common regulatory and legal framework by introducing the concept of ‘unacceptable’. AI has been classified based on risk categories and accordingly the legal intervention. It has a third-party assessment.

“India has the concept of data protection but questions encryption by WhatsApp, the New Digital Act is not covering the grievances of intermediaries where the innovations, and use cases of AI will come whether it is publishers’ content being used by Meta Data (India is only country that has not defined bargaining code as yet causing injury to publishers ) and need to define bargaining code or the facilitation to gaming, it stops of becoming an umbrella act with the opportunity to allow innovations and use of AI,” said Aruna Sharma, Former Secretary, Ministry of Electronics and Information Technology, GoI.

Sharma said that the manufacturing is going to be AI-based, so the health diagnosis and education scoring, the delay, wait and watch policy is not going to work. The need is for India to become a hub as we have technical manpower but an umbrella Digital Act is the need of the day that is flexible to give quick relief.

AI and its challenges need a framework that recognizes the changing paradigm. Every country needs more thoughtful regulation that can balance innovation with risks and harms.

“India has had two advisories, but both lack thoughtful inputs.India has been notoriously bad at protecting its citizens from data theft, surveillance, and cyber frauds. No amount of regulations can help if enforcement and implementation are lacking,” said Mishi Choudhary, Founder, SFLC.in (Software Freedom Law Center, India).

The EU’s new AI Act, a global first, regulates AI not just within the EU but also applies to companies outside the EU whose AI affects EU citizens or uses their data. With robust regulations emerging in various jurisdictions that could affect tech companies operating out of India, it is crucial for India to stay competitive. “We need a legal framework that tackles risks like liability for AI-caused damage and ensures responsible innovation. Whether through a dedicated AI Law or integrating AI regulations into existing sectoral laws, India must choose the best path to ensure safe and responsible AI development that fosters innovation. In the meanwhile, an “AI regulatory sandbox” could allow for testing regulations in a safe space,” said Dr Amar Patnaik, former member of the Parliament, Rajya Sabha.