Zarlish Fatima

If recent global discussions on Artificial Intelligence (AI) policy have taught us anything, it is the regulation of artificial intelligence. Recent global discussions on AI policy underscore the critical need for artificial intelligence regulation. AI has emerged as the pressing issue confronting policymakers in the 21st century. The widespread deployment of AI has revealed critical weaknesses in existing regulatory frameworks. This increasing use of AI has exposed critical weaknesses in existing regulatory frameworks, including the lack of transparency, human oversight, weak data protection laws and poor enforcement. As the OECD (2019) highlights, much of this stems from insufficient monitoring and the absence of clear accountability structures for how AI systems operate.

The public’s confidence in democratic, legal, and regulatory institutions has been compromised during the time of this technological revolution. The UN Secretary-General claims that “trust in global institutions is at breaking point”, citing persistent accusations of unfair treatment and broken promises (UN Secretary-General António Guterres, Address to the General Assembly, 2023). These legitimacy crises have emerged alongside the growing use of AI tools in public life, such as:

●      Facial recognition in policing,

●      Algorithm-driven content on social media

●      Automated decision-making in welfare systems.

The governments of India, Pakistan, and Sri Lanka have launched ambitious plans for digital transformations. Yet AI laws are still fragmented and inadequate. This regulatory gap increases the risks of algorithmic opacity, where decision-making processes are hidden and unaccountable. According to UNESCO (2021), this would exacerbate inequality and undermine human rights in societies that are already grappling with socio-economic disparities.

The evidence is both clear and deeply disturbing. Multiple state governments in India have used AI-powered systems. For instance, AI-powered facial recognition technology for law enforcement. But often used in the absence of transparent legal safeguards or impartial oversight. Human rights groups have cautioned that such deployments could strengthen the discriminatory practices against marginalised communities (UNESCO, 2021). Pakistan has no specific legislation for controlling data protection or algorithmic accountability. Although the Ministry of Information Technology and Telecommunication has a draft of the National AI policy, it is still in the consultative stage (Government of Pakistan, 2022).

Policymakers worldwide have begun to respond to the ethical and legal dilemmas posed by AI. The European Union's planned Artificial Intelligence Act will introduce a risk-based regulatory framework that imposes stringent obligations on high-risk systems (European Commission,2021).

Around the world, legislators are starting to address the moral and legal conundrums raised by AI.  High-risk systems must adhere to strict requirements under a risk-based regulatory model. It was introduced by the European Union’s proposal of the Artificial Intelligence Act. Moreover, the OECD has articulated the principles in 2019 that demand for, transparency, accountability and human-centred designs.

Furthermore, South Asian countries are in the early stages of translating these principles into regulations and enforcement. The repercussions of inaction are clear. According to the Stanford Institute for Human-Centred Intelligence study in 2023, more than 40% of AI systems deployed in public policy situations lacked effective audit or redress. These risk factors are amplified in South Asia. In regions, institutional capacity and public awareness are often limited. Without transparent governance, these AI-enabled tools and systems could be misused. For example, these can be misused for:

1. Political surveillance
2. Undermine electoral integrity
3. Exacerbate biases in Welfare Distribution.

Historical records can teach an important lesson. UNESCO (2021) has emphasized previous technological advancements, nuclear energy, and industrial manufacturing revolutions. These demonstrate that when innovation surpasses regulation, societies grapple with harms that can endure for generations. The notion that technological superiority can bring about wealth while sidestepping ethical and legal obligations remains dangerously misguided.

Suppose governments, civil societies, and businesses in South Asia don't act quickly to draft a strong framework. In that case, AI's promise may give way to a norm defined by discrimination, lack of transparency, and systematic risk. OECD emphasized that the trustworthy adoption of AI will not happen without a solid foundation of shared and clear rules.

The message is obvious: Governance cannot be a secondary priority. It must build the foundation for the region’s Digital future.

Disclaimer: The views expressed in this article are solely those of the author and do not necessarily reflect the official stance of The Himalayan Research Institute Pakistan - (THRIP)

_________________________________

Zarlish Fatima is a political Science student at Kinnaird College, a youth advocate, and the founder of Soch.Sawal is a platform for critical dialogue on governance, rights, and digital futures in South Asia.