LocationsA Competition with Time
By Ammara Durrani | Assistant Resident Representative, UNDP Pakistan
AI is propelling ‘digital’ by leaps and bounds through a simple metric – transitioning from the era of ‘the internet of things’ to the ‘speed of things’. Pakistan cannot fully embark on AI’s high-speed highway until the country graduates from its current Access-Adoption phase.
Back to the Future
At UNDP, we strive for the Pakistan National Human Development Report (NHDR) to serve as premier evidence and thought leadership framework for development policy debate and practice. In 2022, coming out of the COVID-19 pandemic and in a fast-changing country context, we had several development challenges to consider. In everything we analyzed as a potential subject for our next NHDR, the phenomenon of ‘digital’ jumped at us. The all-pervasive presence and power of ‘digital’ in every national sector of Pakistan was obvious. But we also noted an absence of a normative public policy approach, in contrast to the digital industry’s rapid growth.
UNDP’s global Strategic Plan 2022-2025 envisioned scaling up the global development gains by propelling three specific Directions of Change: a) Structural transformation for green, inclusive and digital transitions; b) Leaving no one behind by centring human agency and development; and c) Building resilience to respond to systemic vulnerability and risk. Surviving the COVID-19 and learning from new development approaches applied during the pandemic years of 2019-2022, the Plan combined these three directions of intended change with three powerful Enablers: a) Strategic innovation; b) Digitalization; and c) Development finance.
We adopted digital transformation to become a better development organization. We then applied our knowledge in countries to help them experience better development. In Pakistan, one of our key takeaways from the pandemic was that in the post-COVID-19 era, a new digital social contract emerged to find solutions that reshape and improve traditional service delivery through technology, reaching millions. In fact, UNDP helped to adopt 580 digital solutions in 82 countries (including Pakistan) in response to the COVID-19 pandemic period of 2020-2021, including 96 data collection systems, 71 e-commerce systems, 61 e-learning platforms, and 149 e-governance systems.
All of this led to the milestone publication of the Pakistan NHDR 2024 entitled “Doing Digital for Development: 4 As of Access, Adopt, Accelerate, Anticipate”. As April 2025 marks its first anniversary, it is a fitting occasion to revisit our original research question: What does ‘digital’ really mean as a national whole for Pakistan’s society, government, businesses, international partners, and peoples?
This question was our north star to formulate a trinity of hypotheses: a) Countries’ digital transformation and sustainable development have become inherently unavoidable, mutually inclusive, and increasingly co-dependent; b) Pakistan needs to ‘do digital’ right, better, and faster in order to ‘develop’; and c) By ‘doing digital for development’, Pakistan can improve its Human Development Index (HDI) indicators and outcomes for achieving inclusive, resilient, and sustainable growth.
Over the last two years – and while we are still grappling with the challenge of mainstreaming ‘digital’ and bridging its vast divides in Pakistan – the global thought and policy focus has fast forwarded to Artificial Intelligence (AI) as the newest frontier for countries to achieve innovation, enterprise, and growth.
Speed vs. Stuck
Today, the question is: How can Pakistan ‘do AI’ right and better for national development?
The UNDP NHDR 2024 showed that the existing gaps in digital access and adoption in Pakistan prevent the country from fully leveraging technology for its human development and prosperity. A country of 250 million strong, it has 87.35 million internet users, while internet penetration stands at 36.7 percent. In contrast, 150.8 million (63.3 percent) people in Pakistan do not use the internet. At this level of digital access, Pakistan is home to 71.7 million social media users, comprising 30.1 percent of the total population. There are 191.8 million (80.5 percent of the total population) active cellular connections in Pakistan.
A comparative analysis of leading global indices measuring countries’ digital progress shows Pakistan lagging behind. The country ranked 45th out of 52 countries in the World Internet Development Index (2023) released by China, covering six matrices: information technology infrastructure, digital technology and innovation capability, digital economy, digital government, cyber security, and cyberspace international management. Pakistan ranked 79th out of 120 countries on Inclusive Internet Index covering areas including availability, affordability, relevance, and readiness.
In the E-Governance Development Index covering domains related to telecom infrastructure, human capital, and online services, Pakistan ranks 153rd out of 193 countries. On the Digital Quality of Life Index 2022, it ranks 96th out of 117 countries on key indicators of internet affordability, internet quality, electronic infrastructure, electronic security, and electronic government. Pakistan ranked 97th out of 113 countries on the Asian Index of Digital Entrepreneurship Systems (2021) that captures countries’ general, systemic, and digital frameworks conditions for entrepreneurship.
Overall, Pakistan’s rankings on these indices demonstrate major gaps and weaknesses in digital accessibility, availability, and relevance; though, it fares better on affordability.
A look at Google’s generative AI window on “AI in Pakistan” shows the glass half empty (see Box). On the policy side, the narrative seems merely to have been updated by largely replacing the word ‘digital’ with ‘AI’. Institutional approaches and policy responses reported in the media thus far do not give evidence of a focused sectoral approach, a bench-marked growth trajectory, or a clear matrix of responsibilities. The government has begun annual budgetary allocations for AI-related projects, but expenditure and impact trails are unreported and unanalyzed. The Ministry of Information Technology’s 2024-25 budgetary allocation was Rs. 24 billion ($85.5 million) compared to Rs. 723 million ($2.5 million) allocations for AI in the 2022-2023 budget. The glossy idea of AI has caught policymakers’ attention and fiscal patronage, but a clear action plan and experienced leadership are absent.
This is quite like how the digital sector evolved and grew in Pakistan – organically, highly unevenly, without policy coherence, and without a targeted national action plan. Successes continue to be led by the private sector and civil society, while government remains a runner-up.
Statista’s data from March 2024 projected Pakistan’s AI market size to reach US$861 million in 2025, compared to the U.S. market size projection of US$74 billion in 2025. Pakistan’s market size is expected to show an annual growth rate of 26.28 percent, resulting in a market volume of US$3.49 billion by 2031. The country’s growing use of AI is driven by a shift towards more efficient and cost-effective business operations. There is a surge in demand for AI-powered chatbots and virtual assistants, with businesses looking to automate customer service and improve efficiency. Additionally, there is a growing trend of using AI for predictive analytics and data-driven decisionmaking in healthcare, finance, and agriculture.
Today, our experiential realization is that AI propels ‘digital’ by leaps and bounds through a simple metric – transitioning from the era of ‘the internet of things’ to the ‘speed of things’. But Pakistan cannot embark on AI’s high-speed highway to accrue national benefits at scale until the country graduates from its current Access-Adoption phase. AI is transforming economies and societies (mostly in the Global North) that are already in a mass state of digital readiness. Though Pakistan’s neighbors, China and India, have emerged as examples of regional AI readiness and competitiveness.
Understand the AI Balance of Power
American technology company OpenAI was founded in December 2015. Its November 2022 debut of ChatGPT, an AI chatbot built on top of OpenAI's foundational large language models (LLMs) like GPT-4 and its predecessors, succeeded in projecting artificial general intelligence (AGI) as a pop culture phenomenon in the realm of digital. Within five days, the chatbot had attracted over one million users. OpenAI defines AGI as "highly autonomous systems that outperform humans at most economically valuable work" and that also, quite ironically, “benefits all of humanity”.
In January 2025, China unveiled the rival, economically cheaper, and technologically more efficient DeepSeek-R1, which has arguably shifted the AI balance of power towards Asia. The launch of DeepSeek-R1 took Silicon Valley by surprise. The U.S. AI hardware leader Nvidia's share price dropped sharply, losing US$600 billion in market value, the largest single-company decline in U.S. stock market history. The company DeepSeek reportedly recruits AI researchers from top Chinese universities as well as from outside Information Technology fields to broaden its models' knowledge and capabilities.
This growing U.S.-China rivalry over AI is also converging with the tsunami of trade tariffs competition between the two countries (and beyond) that has risen this year following the return of President Donald Trump to the White House. The Time magazine reported on April 8th that stocks in AI companies were among the biggest losers after President Trump announced sweeping tariffs on foreign trading partners in the preceding week. Owing to the AI industry’s import reliance on overseas supply chains for chips and datacenter equipment, tariffs will increase its gigantic costs that are already running in billions of dollars.
As a leader of the Global South, India is also recalibrating its AI approach and footprint. There is increased policy rhetoric for India to compete with the U.S. and China, rather than risk becoming their ‘digital colony’ and using their AI systems for achieving India’s targets of reaching 10 percent GDP growth by 2023 and creating a $1 trillion digital economy.
Beyond geopolitical and geoeconomic ambitions, India is also focusing on developing a homegrown AI economic model of ‘startup capitalism’ by equipping local firms with the necessary financial, infrastructural, and data access resources to drive national innovation and competitiveness. This approach also mirrors the sectoral and specialized ‘Anticipate-Accelerate’ model that the NHDR 2024 has already prescribed for Pakistan.
Prioritize a Capacity-Content Model Over Competition
With a ‘low’ HDI ranking of 164th out of 193 and under-developed human capital levels (comparable to Sub-Saharan Africa), Pakistan’s best bet to avail AI’s advantages will be to prioritize the building of human and skills competitiveness and locally relevant content that can ultimately position the country for AI readiness.
A 2024 World Bank study entitled “Global Trends in AI Governance: Evolving Country Approaches” provides a very useful policy framework for a country like Pakistan that still needs to switch from the ‘Access-Adopt’ lane into the ‘Anticipate-Accelerate’ frontier territory. The World Bank framework prioritizes five factors for policy interventions that are highly relevant for Pakistan:
Policy Priorities and Local Context: Focus on promoting trust, protecting fundamental rights of users, increasing digital inclusion, fostering local innovation ecosystems, increasing market competition, and attracting investment.
Maturity of AI Ecosystem: Focus on development and availability of infrastructure and human expertise, promoting good data governance, introducing baseline governance requirements, early binding measures to prevent AI risks, and harmonization with global standards.
Fahad Maqsood Qazi working on the artificial intelligence dubbing system in Pakistan’s southern city of Hyderabad on April 5, 2025.
Legal Framework and Regulatory Environment: Focus on AI-specific provisions in data protection, cybercrime, competition, human rights law, and dispute resolution mechanisms.
Public Resources and Capacity: Use available public resources for binding measures (more investment and skilled staff with lasting oversight mechanisms) as well as self-governance and soft law approaches (less investment with some centralized public monitoring); and
Stakeholder Ecosystem: Include government bodies, industry participants, academia, civil society, and consumers. Ensure market trust and active participation.
On April 5th, Arab News highlighted the story of 23-year-old Fahad Maqsood Qazi from Hyderabad, Sindh who has developed previously unavailable Sindhi text-to-speech and speech-to-text AI models and has shared on open-source platforms. Fahad has bridged the AI gap for a language spoken by around 40 million people globally, including Pakistan and India.
Ignited by their self-taught passion, skills, and knowledge, let a thousand Mahroses and Fahads bloom in Pakistan for the country to meet and match AI in real time.