Like any other reading available online, I would not be saying AI has revolutionised our lives and that there will not be any labour-intensive jobs available anymore. I challenge this prevailing approach where public discourse surrounding artificial intelligence has been blown out of proportion, prompting people to embrace it due to social influence, peer pressure and the fear of being left behind.
Everett Rogers argued that innovations spread through social systems because people imitate one another and follow trends. This social attitude leads to the rapid adoption of new technologies, often outpacing AI literacy and formal training. Pakistan mirrors this trend. As per a survey of 2025, 86 per cent of professionals use AI at work, while only about half of them have actually received any formal training. So, Pakistan is rapidly adopting AI, but that rapid adoption should not be confused with technological progress.
My question is that everyone uses ChatGPT, companies automate customer support, and universities introduce AI courses, but does any of this mean that Pakistan can build foundation models, AI chips, and research laboratories? Pakistan is investing in the adoption of AI solutions while neglecting the institutions that actually produce innovation.
The rapid adoption of artificial intelligence in Pakistan is consistent with the Technology Acceptance Model, where users embrace technology when they perceive it useful and easy to use. This adoptability is highly characterised by consumption rather than technological capability. In the case of Pakistan, AI adoption is on a fast track but without AI literacy. For instance, ChatGPT is often used as a shortcut rather than a cognitive partner. These practices only contribute to cognitive offloading and consequently fail to encourage meaningful investment in developing technological capabilities.
Pakistan’s AI problem is that it is prioritising AI consumption over AI creation; the country is producing AI consumers rather than AI creators. Such an investment model may deliver immediate efficiency gains, but without strengthening domestic technological capabilities. The economic value of artificial intelligence is captured by those who own the models, chips, data and platforms, not those who merely use them. Thus, the country will remain at the consumer end of the global AI value chain instead of becoming a producer of homegrown technologies.
According to Cohen and Levinthal’s concept of absorptive capacity, a country cannot benefit from new technology unless it has the capacity to recognise, understand, adapt and develop it further. This suggests Pakistan must invest in research institutions, human capital and knowledge systems leading to AI innovators.
As of today, it may not sound as urgent or may seem like a distant concern, but in the near future, a new strategic necessity will be AI sovereignty. Throughout history, nations have competed for food and water, military strength and nuclear weapons; the coming decades will increasingly be shaped by competition over AI infrastructure. Freelancing and service-based startups should not be confused with deep-tech innovation. The country’s startup ecosystem lacks funding, not talent. Pakistan possesses significant AI talent potential, yet it lacks the political will and investment to capitalise on this human resource.
The latest P@SHA Skills Report highlights a workforce that is abundant in numbers but uneven in AI readiness, further pointing to a mismatch between industry demand and academia. This is also one of the reasons for brain drain, as people move abroad in search of better opportunities.
When a country fails to develop a research-conducive environment, it pushes away its talent. For Pakistan, the challenge is not just the availability of financial resources but the investment priorities. An investment, even though small in amount, if not made in the right direction, would not produce meaningful results. China has treated artificial intelligence as a strategic asset and heavily invested in the right direction — that is, research institutions and infrastructure development.
The distinction here is that encouraging the use of AI is not wrong, but a long-term strategy should not be focused on it solely; rather, it should include AI as a part of long-term AI self-sufficiency. To achieve that goal, Pakistan needs to strictly adhere to its National Artificial Intelligence Policy 2025. Moreover, there is a need to build a system where public-private partnerships would encourage indigenous AI models.
The government can become the first consumer of startups to promote their innovation activities. A sophisticated, state-of-the-art research laboratory must be established in each province where future researchers can be nurtured. So, to increase investment and become a regional leader, we need to move from AI consumption to AI capacity building. Without this shift, no investment will produce sustainable results.