Published in May-Jun 2023
Pakistan recently unveiled its National Artificial Intelligence Policy. Spearheaded by the Ministry of Information Technology & Telecommunications (MoITT), the policy is an exhaustive document that delineates the objectives, key drivers and cornerstones of an AI-driven society, establishing a balanced blueprint of goals and targets.
This said, the policy appears to fall short in several critical areas, most notably its disregard for the private sector and the potential benefits the sector could bring to the AI landscape. The policy seems anchored in the Public Sector Development Programme (PSDP) funding mechanisms, with IGNITE funding projected to absorb 30% of the budget – an approach that severely curtails the policy’s scope, thereby compromising its capacity to shape the future of AI regulation, research and management in Pakistan.
Although the policy suggests international funding or grants as a viable financial resource, it fails to account for the limitations inherent in such funding mechanisms. Often, international funds come with a set of obligations that may impede the range of research, potentially stifling the generation of new knowledge. The policy’s proposition to establish AI Centres of Excellence in key cities, including Karachi, Lahore and Islamabad, is ambitious, yet it does not convincingly address the need to stimulate demand for innovative AI research. It lacks a creative approach to capitalise on AI as a tool for fostering new industry opportunities and propelling cutting-edge research that could bolster Pakistani businesses in the global supply chain.
The strategy also glosses over the role of the Internet of Things (IoT), hinting at a shortfall in its vision of IT as an integral pillar of an AI ecosystem. As we forge ahead towards an AI-enabled society, the questions of privacy and data usage become crucial as such data could reveal societal patterns, facilitating better governance and policy guidelines for improved living standards. IoT needs to become a mass production data-producing mechanism and play a central importance in the policy framework.
A glaring omission in the policy is the lack of attention given to the potential of AI-enabled agricultural technology (AgriTech). AI can vastly enhance land productivity and farming skills, while improving soil and seed quality. This transformation could require collaborations between various institutions of different faculties and expertise to foster new knowledge as well as acceptance among the communities engaged in the agricultural supply chain, thereby opening up new AgriTech opportunities.
From a governance point of view, the delivery of services needs to transition to a system-centric approach, with AI identifying and rectifying system faults. This shift necessitates a reorientation of service delivery institutions in the short- to mid-term, potentially influencing social job structures. The policy also needs to work with other government institutions to create harmony within the system for the acceptance and implementation of these changes.
The policy’s approach to promoting AI awareness and knowledge appears to be wedded to traditional methods, such as public awareness messages via the Ministry of Information and Broadcasting and the Pakistan Telecommunication Authority. This approach overlooks changing media consumption habits, especially among younger demographics who are more likely to engage with digital and social platforms. Yet, it is essential to capitalise on these trends to ensure effective communication and awareness about AI’s opportunities. This could involve collaborations with influencers, the creation of engaging digital content and interactive online forums to discuss AI implications and advancements.
The proposed model for upscaling and reskilling seems to cling to an outdated format of physical training centres. The evolution of AI will bring forth personalised training according to individual needs rendering mass-focused training models increasingly obsolete. The future of AI lies in its ability to assist individuals at their convenience, and according to their skillsets. The education and skills development systems need to evolve so as to incorporate these changes and foster an environment of continuous learning and adaptation. The policy seems to ignore this aspect.
The proposed strategy for fostering fresh talent and establishing internships focused on AI seems to be dictated heavily by government control. This approach may not be efficient in creating a practical system for nurturing AI-empowered citizens. Given the demographic realities of a country like Pakistan, where a large segment of the population is young, the proposal to create a limited number of AI internships seems insufficient. In a country that is churning 150,000 people per month as 18-year-old young adults, a mere 25,000 AI internships annually seems not only inappropriate but inefficient. A better approach could involve broader collaboration between private sectors, academia and the government to create a larger number of opportunities and facilitate the transition of the government’s role from controlling to enabling.
The policy indicates an understanding of the importance of data privacy and protection, which is a critical component of an AI-driven society. However, it does not sufficiently address the transformation that is required in attitudes toward data governance. As we move towards an era where citizens should have the right to data sovereignty, the role of government as the gatekeeper of data will need to be redefined. This will require a significant shift in Pakistan’s security apparatus and privacy parameters to ensure data democratisation.
While the policy demonstrates a keen interest in developing AI-focused curricula, it overlooks the dynamic nature of this evolving field of study with AI-enablement. With real-time updates and emergent curricula, the roles of education and technology boards may require significant redefinition and reorientation. This may require policy transformation in the regulatory bodies for education and the content innovation taught at institutes of learning.
The drive for industrial transformation through AI should be directed by incentivising industries to adopt AI-enabled procedures and systems. This could be achieved by implementing IoT products and services, thereby promoting and incentivising data generation. In an AI-enabled society, data generation by every aspect and player in society is crucial to ensure digital justice and opportunities for all.
The National Artificial Intelligence fund, while a positive step towards promoting AI awareness and innovation, should not be the only mechanism to facilitate growth. Thought leaders should consider sector-specific approaches that can complement societal vectors, thereby creating new jobs and opportunities and defining new vectors to create value around new breakthroughs.
In conclusion, while the National Artificial Intelligence Policy is a commendable initiative toward becoming an AI-centric society, it requires revisions and additions to ensure inclusivity, creativity and efficiency in the AI landscape. More importantly, it should serve as a catalyst for social transformation and progress, leveraging AI to propel Pakistan into a prosperous future. MoITT needs to invite more input from the social scientists in addition to the people managing the technical side of AI developments.
Amir Jahangir is a global competitiveness, risk, and development expert. He leads Mishal Pakistan, the country partner institute of the Centre for the New Economy and Society Platform at the World Economic Forum. firstname.lastname@example.org