Regulation of AI – Opinion News

Written by Sandeep Parekh

Development of artificial intelligence is progressing rapidly AI is already transforming industries and disciplines, with the European Union (EU) being the first jurisdiction to introduce legislation to regulate AI. Broadly speaking, AI that poses an unacceptable risk will be banned, while AI systems that pose minimal risk will not be regulated. The former category includes AI that affects people's rights, biometric classification systems, scraping facial images from sources to create facial recognition databases, social scoring, predictive policing, and manipulating human behavior. and AI that exploits people's vulnerabilities. Additionally, the law establishes the EU AI Authority, which will be the nodal body for the implementation and enforcement of AI laws. AI law, like general data protection regulation, has an element of extraterritorial jurisdiction.

What is actually regulated is AI that is not classified as either unacceptable or minimal risk. These are high risk AI systems and limited risk AI systems. High-risk AI systems pose a threat as they include deployments to critical infrastructure, education, essential services, law enforcement, justice enforcement, governance, and more. Such systems are required, inter alia, to be registered, evaluated and evaluated in the relevant EU databases. Reduce risk and ensure transparency, accuracy, and more importantly, human oversight. Additionally, people will have the right to file complaints about AI systems and be held accountable for decisions related to high-risk AI systems that impact the rights of victims.

stock market AI penetration is not the exception, or the exception. While recent breakthroughs have primarily been in the field of generative AI, big steps are being taken to give AI more power and expand the data it has access to.

Until 30 years ago, most activities related to securities market trading, such as research and orders, incorporated little technology.. The focus on the use of technology began in some earnest after the introduction of dematerialized stocks.From then until now, India is leading the implementation of the T+0 payment cycle.

With the introduction of AI, the securities market will undergo a new transformation. However, one concern related to this is around data privacy. As AI and AI-generated algorithms permeate various sectors, including securities markets, regulators face the challenge of enacting laws to effectively govern these technologies.

Algo trading and robo advisory

Currently, algo trading is defined as a trade executed through automated means. Recent proposals by Sebi Regulating algo trading has received mixed reviews, and the regulator's approach to achieving that objective has faced some criticism. It is important to remember that AI can effectively write code based on the instructions it is given. Therefore, in the near future, creating algos may no longer be the exclusive domain of trained IT professionals. If the introduction of an algorithm created by AI violates securities laws, the question is to what extent the person who created the algorithm using AI will be held liable. There is a principle that the developers of AI, the humans behind the “machines”, are responsible. However, advances in AI are unpredictable at this stage and may need to be reconsidered. As AI changes the landscape around us, our laws need to keep pace, so that the rights and obligations of those involved are predetermined.

Additionally, robo-advisors will likely take center stage in the distant future. For example, with vast amounts of data points being analyzed in seconds and investment strategies created in seconds rather than weeks, it's not unreasonable to assume that there has been a major shift in the way investment advisory services are performed today. , a review of the existing regulatory framework may be warranted, requiring both regulation to be strengthened and streamlined.

Complaint redress and enforcement

When properly (and securely) integrated into judicial systems, AI could eventually be used to administer justice. In fact, it has recently been suggested that AI could be used to resolve minor traffic challans once such capabilities are properly built. Similarly, securities regulators may consider initiating the process of developing AI that can effectively monitor, supervise, and assist in the enforcement of securities laws.

Additionally, the recent focus on online alternative dispute resolution mechanisms in the securities markets may create minor challenges depending on complexity and amount./assets, and depending on the nature of the dispute, the AI ​​acts as an arbitrator or mediator.

Pattern recognition and predictive analysis

What could be a game-changer for regulators is the development of AI models that are increasingly efficient at recognizing patterns and understanding the data points that AI has access to and how they are “coded” to “think.” ” is to predict the “future'' depending on what is being said. . Algos are already being deployed by regulators around the world to identify and/or track suspicious activity, and AI can be of invaluable help in this regard. For example, Sebi recently issued a circular introducing the use of blockchain to verify information and ensure transparency between intermediaries and entities. Integrating AI into such systems can help predict defaults and prevent violations, protecting investor interests. However, such technology must be used with caution, and stringent safeguards must be built around such systems to prevent misuse.

Therefore, while the introduction of AI in securities markets will lead to increased efficiency, reduced costs, and enhanced decision-making capabilities for market participants, it will also raise significant risks regarding market manipulation, algorithmic bias, data privacy, and systemic risks. Concerns arise. The need for regulatory oversight and a comprehensive legal framework to address issues arising from the use of AI. As AI continues to evolve and reshape the securities trading landscape, regulators will remain vigilant, taking an adaptive and forward-thinking approach to balance innovation and regulation, thereby ensuring that innovation and regulation emerge from the intersection of AI and regulation. We need to overcome complexity. Stock market. It is better to delay introducing too many regulations until the problems on the ground have subsided.

The author is a managing partner at Finsec Law Advisors. Co-authored with his Parker Karia, Senior Associate at Finsec Law Advisors.

Disclaimer: The views expressed are personal and do not reflect the official position or policy of Financial Express Online. Reproduction of this content without permission is prohibited.