Legal Advice on Artificial Intelligence law in Pakistan

Embracing the Future: Pakistan’s National AI Policy

In the fast-evolving landscape of technological advancements, one particular force is reshaping the world as we know it – Artificial Intelligence (AI). Recognizing the transformative potential of this cutting-edge technology, countries worldwide are racing to harness its power, and Pakistan is no exception. With the recent launch of the National AI Policy by the Ministry of Information Technology & Telecom (MoITT), Pakistan has taken a significant step towards embracing the possibilities of AI and ensuring national competitiveness and citizen welfare.

The comprehensive 41-page document lays out a strategic roadmap with the aim of making Pakistan a technology-empowered nation. The policy focuses on several key areas, including creating public awareness about AI, developing a skilled workforce, investing in research and development, establishing a national AI fund, and improving infrastructure for AI-related technologies. It also emphasizes launching national AI projects to address social and civic issues, promoting responsible AI usage, developing regulatory mechanisms, and ensuring accessibility to AI models. Additionally, the policy seeks to revolutionize the public sector, provide sectorial support, and empower academia with AI technology.

The pillars on which the policy stands are Market Enablement, Creating AI Awareness & Readiness, Building a Progressive & Trusted Environment, and Transformation and Evolution. These pillars form the foundation for the integration of AI across various sectors, with special focus on agriculture and manufacturing, which are crucial components of Pakistan’s economic landscape.

One of the policy’s commendable aspects is its emphasis on inclusivity, particularly in regards to women. By prioritizing the involvement of women in the AI journey, the policy seeks to create a more equitable and diverse AI ecosystem in the country.

However, despite the positive strides made with the launch of the National AI Policy, numerous challenges must be addressed to ensure its successful implementation. One major concern is the lack of specific data laws, raising questions about the policy’s credibility. To ensure a cohesive effort, it is crucial for relevant ministries like Defence and Science and Technology to actively participate in the policy implementation process.

Furthermore, Pakistan faces a shortage of adequately trained AI professionals. To bridge this gap, the policy ambitiously aims to train a million IT graduates in AI and related technologies by 2027. However, this task requires robust efforts and commitment to counter the challenge of ‘brain drain,’ where talented individuals seek opportunities abroad.

Additionally, the absence of sufficient infrastructure in rural areas poses obstacles to AI deployment. Public-Private Partnerships (PPPs) are promoted in the policy to address this, but their complexities must be navigated carefully.

Notably, the policy does not fully explore AI’s potential role in the defense sector, despite its importance in the global military landscape. Moreover, the review section of the policy’s implementation raises concerns about its practicality and long-term effectiveness.

To overcome these challenges and fully leverage the potential of AI, Pakistan must adopt a multistakeholder approach involving the government, industry, academia, and civil society. Implementing the policy’s objectives will require a visionary and persistent effort, transcending political cycles.

As of 2023 Pakistan stands at the crossroads of unprecedented opportunities and formidable challenges in its journey towards AI integration. With sincere commitment and collaborative efforts from all stakeholders, Pakistan can pave the way for a safer, more responsible, and pragmatic AI-powered future. By treading this path with determination, the nation can unlock the socio-economic benefits of this innovative frontier, ultimately transforming lives for the better.

In the era of exponential technological advancements, Artificial Intelligence (AI) stands out as a transformative force that holds the potential to reshape our world. Governments worldwide are increasingly recognizing the critical need to address the challenges and opportunities presented by AI to ensure their survival and progress. The field of AI is rapidly evolving, with algorithms capable of self-learning and automation of complex tasks. As AI continues to advance, it has significant implications for various sectors, including defense and governance.

The integration of AI technologies in defense organizations is essential for achieving Network Centric Warfare (NCW) capabilities. NCW enables information technology to turn information superiority into a competitive advantage. However, today’s major advances in Information Technology are primarily driven by the demands of the commercial sector, leading to disruptive technologies like big data analytics, AI, and IoT infrastructure. These technologies are redefining how governance and operations take place within System of Systems (SoS) in defense organizations.

The challenges and opportunities posed by AI in governance are substantial. The uncontrolled growth of AI in military robotics raises concerns about proper governance to prevent potential catastrophic consequences. Data deluge and the proliferation of IoT networks further complicate the management of massive volumes of data generated in military operations. For effective governance, government executives, military strategists, and the business community must proactively engage in AI strategy formulation and risk identification. The development of AI expertise and integration with core defense functions and services is critical for effective implementation.

IoT infrastructure, including smart metering, battlefield monitoring, unmanned systems, and more, plays a crucial role in modern defense. To ensure independence and safeguard against foreign influences, an indigenous data analytics and scenario building futures laboratory is required. Implementing a combat cloud infrastructure will enable greater operational flexibility and joint operations. Lead System Integrators (LSIs) can help manage complex technology projects in defense, considering the sector’s challenges and pressures.

The impact of AI is not limited to the defense sector but extends to all industries, including traditional fields like the Indian IT industry. As AI continues to transform industries, governments need leaders well-versed in technology and proficient in technology governance. Soldiers in the battlefield will rely on AI and IoT-enabled components, making strategic and tactical decisions in an information-rich environment. Technology governance with AI presents unique complexities that require careful consideration for the future of national security.

Advanced countries are already making strides in AI governance, setting the foundation for domestic policymaking on machine intelligence. Technology governance with AI requires a new approach, as thinking machines can make decisions independently. The focus must be on informing the new national security paradigm and preparing for a future where AI plays a crucial role in defense and governance.

Pakistan, too, must embrace AI-governance to harness the potential of AI while safeguarding against potential risks. Developing an AI-ready workforce, investing in indigenous data analytics, and integrating AI into defense functions are essential steps for Pakistan’s future. By adopting a proactive approach to AI-governance and fostering collaboration between different stakeholders, Pakistan can position itself as a competitive player in the AI-driven world and ensure its future growth and security.

Exploring Pakistan’s Draft Artificial Intelligence Policy: The pathway to a Knowledge-Based Economy

In its unwavering pursuit of the Digital Pakistan vision, the Ministry of IT & Telecom is poised to lead Pakistan into an era of innovation and development. At the heart of this transformation lies the Draft Artificial Intelligence (AI) Policy, a visionary roadmap that seeks to establish a conducive ecosystem for AI adoption while ensuring responsible and equitable use of this cutting-edge technology.

1. Embracing AI for Good

The AI Policy represents a pivotal milestone in Pakistan’s journey towards becoming a knowledge-based economy. It sets forth a national strategy aimed at harnessing AI’s potential to empower diverse user journeys across various market horizontals and industry verticals. The guiding principle of this policy is to adopt AI efficiently and responsibly, aligning it with the “AI for good” initiative by the International Telecommunication Union and the Sustainable Development Goals of the United Nations.

Acknowledging the transformative nature of AI, the policy advocates a cautious approach that rethinks AI adoption in the local context, keeping in mind the relevance of the existing job market and the needs of the growing population.

2. Objectives and Attributes

The AI Policy strives to be evidence-based, target-oriented, user-centric, forward-looking, and overarching. It aims to foster a progressive and trusted environment for AI adoption, generating economic gains and improving lives while anticipating and delivering personalized services through responsible AI use.

To achieve these objectives, the policy outlines several key attributes:

2.1 Market Enablement: The establishment of research & innovation centers in AI is emphasized for developing, test-bedding, deploying, and scaling AI solutions. This includes enhancing governance and managing the impact of AI in various sectors.

2.2 Progressive and Trusted Environment: Responsible AI usage is prioritized to boost economic growth and enhance government capabilities for delivering personalized services.

2.3 Enabling AI through Awareness and Readiness: Efforts to increase awareness and understanding of AI technologies and benefits are to be coupled with equipping the workforce with necessary AI competencies.

2.4 Transformation & Evolution: By creating awareness and offering training programs through sectoral cooperation, AI adoption in various industries and sectors is encouraged.

3. Why Pakistan Needs AI Policy

The need for a National AI Policy stems from a holistic approach that addresses multiple facets of AI adoption. These include raising awareness about AI-based platforms while safeguarding privacy, upskilling the workforce on AI and allied technologies, guiding investments in AI research and development, and ensuring ethical and responsible AI use.

Key focus areas of the policy include:

3.1 Investment in Research and Development Infrastructure: Establishing in-house AI capabilities for application development based on local data, including funding research organizations, data standardization, and fostering partnerships with universities and research institutions.

3.2 Ethical and Responsible Use: Proposing the establishment of an AI regulatory directorate to ensure transparent and accountable AI use, free from discrimination.

3.3 Addressing Job Displacement: Investing in training programs to help the workforce adapt to changing job requirements.

4. The State of AI in Pakistan

Pakistan’s burgeoning young population presents both a tremendous opportunity and a responsibility to embrace AI for accelerated socio-economic development. Notable initiatives such as the Presidential Initiative for Artificial Intelligence & Computing and the National Center of Artificial Intelligence have laid the foundation for progress in the field.

However, to achieve tangible impact on Pakistan’s GDP and socio-economic challenges, a comprehensive national strategy for AI adoption is imperative. The AI Policy provides a structured approach to address these challenges and proposes the establishment of a National AI Fund to support crucial interventions.

Pakistan’s Draft Artificial Intelligence Policy represents a bold step toward shaping the nation’s future as a knowledge-based economy. By fostering a responsible and equitable AI ecosystem, Pakistan aims to unlock the potential of AI for the greater good of its people, while also navigating the challenges and opportunities that come with this transformative technology. As the policy takes shape and gains momentum, it holds the promise of revolutionizing various sectors and improving the lives of citizens, driving Pakistan towards a brighter and digitally empowered future.

Pakistan’s Draft AI Policy: Fostering a Hybrid Intelligence Ecosystem for Equitable Progress

Introduction

Pakistan’s Draft AI Policy envisions a future where AI is harnessed responsibly, respecting human intelligence while embracing the transformative power of AI. The policy’s scope is comprehensive, encompassing awareness, skill development, standardization, and ethical use, with strategic objectives divided into developmental and responsibility agendas. Driven by four pillars, the policy sets ambitious yet realistic targets to propel the nation into the age of the fourth industrial revolution.

1. Vision: Embracing AI while Appreciating Human Intelligence

The government envisions a Hybrid Intelligence ecosystem in Pakistan, striking a balance between human ingenuity and AI capabilities. The goal is to utilize AI equitably, responsibly, and transparently to drive sustainable progress and societal development.

2. Scope: A Complete AI-Enabling Ecosystem

The policy framework aims to establish a robust AI-enabling ecosystem in Pakistan. It encompasses all aspects, from creating awareness and skill development to data standardization and ethical usage of AI.

3. Objectives: Development and Responsibility

The policy strategically distributes its objectives across two agendas, emphasizing both development and responsibility:

3.1 Developmental Agenda (Embracing AI)

• Upskilling Human Capital: The policy seeks to train and upskill the workforce in AI to meet market demands efficiently.
• Integrating AI in National Curriculums: To ensure preparedness and application of AI, AI integration into national curriculums is considered essential.
• Promoting Research and Innovation: Encouraging a research-based culture and providing incentives to start-ups and SMEs investing in AI-based technologies and services.

3.2 Responsibility Agenda (Appreciating Human Intelligence)

• Ethical Use of AI: The policy emphasizes the establishment of an AI regulatory directorate to ensure transparent, accountable, and ethical AI use, devoid of discrimination.
• Increasing Public Awareness: Efforts to raise public awareness are prioritized to facilitate sustainable AI adoption.
• Enabling Platform for AI: A sandbox and agile regulatory arrangement will be provided to address societal and regulatory challenges where necessary.

4. Policy Drivers and Targets

The policy identifies four defining pillars to address gaps and challenges, complemented by tangible targets for effective implementation:

4.1 1st Pillar: Enabling AI through Awareness and Readiness

• Raising awareness through proliferation programs about AI’s potential benefits and deployment requirements.
• Improving R&D in AI through applied research and enhancing the footprint of AI through fiscal and logistical support.
• Upskilling the workforce through boot camps and high-impact short courses using MOOCs platforms.

4.2 2nd Pillar: AI Market Enablement

• Addressing societal challenges related to awareness, adaptation, data standardization, and computational needs.

4.3 3rd Pillar: Building Progressive & Trusted Environment

• Fostering a progressive and trusted environment through addressing awareness, adaptation, and ethical challenges.

4.4 4th Pillar: Transformation & Evolution

• Developing roadmaps for sectoral transformation and evolution toward AI adoption, including industrial transformation and sandboxing.

Pakistan’s Draft AI Policy sets an ambitious yet balanced course for the nation’s future. By embracing AI responsibly and appreciating human intelligence, Pakistan seeks to create a Hybrid Intelligence ecosystem for equitable progress. The policy’s comprehensive scope, strategic objectives, and realistic targets demonstrate the government’s commitment to fostering AI’s growth while safeguarding the interests and well-being of its citizens. As the policy unfolds, Pakistan is poised to make significant strides in the era of AI, contributing to global recognition and sustainable development.

An AI  roadmap

Pakistan’s Draft AI Policy lays out a strategic roadmap for embracing AI responsibly while appreciating human intelligence. The policy’s comprehensive scope encompasses awareness, upskilling, market enablement, and building a progressive and trusted environment. To achieve its vision, the policy sets ambitious targets distributed across four pillars, each targeting specific areas for development and responsibility. Let’s explore the policy’s key targets and timeline for implementation.

1. Enabling AI through Awareness & Readiness

Target-1: Public Awareness of AI & Allied Technologies

• Current State: The level of awareness regarding AI and its potential applications is low.
• Desired State: Achieve 90% public awareness of AI and allied technologies among internet-accessible citizens through a comprehensive national program.
• Timeline: By 2026

Target-2: Developing Skilled Human Capital & Upskilling the Existing Workforce

• Current State: Less than 10% of the current workforce in computing and IT is skilled in AI and allied technologies.
• Desired State: Train 1 million new and existing IT graduates in high-impact AI skills and establish 10,000 new trainers.
• Timeline: By 2027

Target-3: Opportunities for On-Job Training for Applied Skillset

• Current State: Limited programs for on-job training of IT professionals are offered.
• Desired State: Introduce incentives for on-job training, targeting 60% of IT graduates annually.
• Timeline: By 2025

Target-4: Higher Education Scholarships

• Current State: Scholarships for AI studies are limited.
• Desired State: Increase AI and allied technology scholarships by 30% annually in partnership with industry and international partners.
• Timeline: By 2026

2. AI Market Enablement

Target-5: National AI Fund

• Current State: No special provisions for Hi-Tech AI initiatives.
• Desired State: Establish the National AI Fund to support Hi-Tech initiatives nationwide.
• Timeline: By 2023

Target-6: Center of Excellence in AI & Allied Technologies

• Current State: Limited regional/central hubs for AI opportunities.
• Desired State: Develop CoEs and Auxiliary Centers nationwide for high-impact skill development, applied R&D, and civic and social initiatives.
• Timeline: CoE-AI by 2026, AuxCenters by 2028

3. Building a Progressive & Trusted Environment

Target-9: Proliferating AI & Allied Technologies Responsibly

• Current State: No provisions for regulating AI.
• Desired State: Establish an AI Directorate under the National Commission for Data Protection to regulate AI responsibly.
• Timeline: By 2024

Target-10: Agile Regulatory Ecosystem

• Current State: Limited AI regulatory sandbox in place.
• Desired State: Establish a regulatory sandbox for AI-led initiatives.
• Timeline: By 2025

4. Transformation & Evolution

Target-13: Transforming the Public Sector through AI & Allied Technologies

• Current State: Limited use of AI in the public sector.
• Desired State: Establish a National/Provincial Data Repository and upgrade existing Data Centers to HCI for AI applications.
• Timeline: By 2024

Target-14: Enabling Academia through AI & Allied Technologies

• Current State: Limited central hub for academia to practice AI models.
• Desired State: Provide access to datasets and computational resources to 150 academic institutions through CoE-AI.
• Timeline: By 2025

Pakistan’s Draft AI Policy seems to demonstrate a well-structured approach to embrace AI responsibly and make significant progress in various sectors. By setting tangible targets and implementing targeted interventions, Pakistan can harness the potential of AI to drive societal development, economic growth, and technological advancement, ensuring a brighter future for the nation and its citizens.

Policy Directives for AI Market Enablement

The policy’s first pillar, AI Market Enablement, aims to stimulate growth in AI by establishing the National Artificial Intelligence Fund (NAIF) and creating Centers of Excellence in AI & Allied Technologies (CoE-AI). Additionally, the policy outlines the implementation of national initiatives to catalyze social development through AI.

1. National Artificial Intelligence Fund (NAIF)

1.1 Allocating Funds: The Ministry of IT & Telecom will direct the Research & Development Fund (Ignite – Technology Fund) to allocate at least 30% of its funds to NAIF perpetually for AI and allied technology research and development.

1.2 Establishment and Administration: The Ministry of IT & Telecom will notify the establishment of NAIF within six months and allocate an initial budget from PSDP funds for the first two years. NAIF will be administered by an independent Board of Directors comprising members from industry, academia, and government, chaired by Secretary/Member IT.

1.3 Funding Sources: NAIF shall have the flexibility to raise funds through international grants, co-investments, partnerships, and incubating startups for early commercialization.

1.4 Auditing and Accountability: NAIF shall be audited annually by a reputable auditing firm with the report made public to ensure transparency and accountability.

1.5 CEO Appointment: NAIF shall hire a Chief Executive Officer from the industry on open merit, responsible for implementing the policy’s directions within the specified timelines.

2. Center of Excellence in AI & Allied Technologies (CoE-AI)

2.1 Demand-Driven Approach: CoE-AI shall focus on a sectorial approach, aligning with industry needs and demands.

2.2 Tiered Model: CoE-AI will have fully-fledged Centers of Excellence in Islamabad, Karachi, and Lahore, and auxiliary centers in other major cities like Peshawar, Quetta, Faisalabad, and Hyderabad.

2.3 Public-Private Partnerships: CoE-AI and auxiliary centers will be established in collaboration with industry stakeholders and operate on a zero-equity model, generating income through services delivered.

2.4 Mandate and Functions: CoEs and auxiliary centers will facilitate demand-driven R&D, incubate startups, develop AI curriculums, provide skill development, offer internships, assist AI regulators, and provide computing infrastructure.

2.5 Performance Evaluation: Monitoring and evaluation experts will regularly assess CoE-AI’s impact and report to NAIF BoD for review and decision-making.

3. Catalyzing Social Development through AI by National Initiatives

3.1 Data Standardization and Aggregation: CoE-AI will orchestrate data standardization in priority social sectors and implement common operating environments for efficient use of AI in public service provisioning.

3.2 Cloud First Policy: National/Provincial IT Boards will host public data on NTC’s cloud infrastructure, ensuring secure and effective service provisioning.

3.3 Smart City Projects: CoE-AI will indigenize computer vision technology for person and object detection to enhance safety in smart city-based projects.

By implementing these policy directives, Pakistan aims to foster a robust AI ecosystem, drive socio-economic development, and leverage AI’s potential for the betterment of its citizens.

4.1.3.2 National Health Services Transformation using AI

This program aims to leverage AI to prevent and better manage chronic diseases, such as diabetes, hypertension, and high blood cholesterol, to reduce the risk of severe complications like heart disease, stroke, and kidney failure. The CoE-AI will collaborate with the Ministry of National Health Services Regulation and Coordination to achieve the following objectives:

• Guidelines for Healthcare Providers: CoE-AI will develop guidelines for healthcare providers to implement AI technology effectively for controlling chronic diseases. AI can analyze clinical data, medical images, health behaviors, and genomic data to create personalized risk scores for individuals. This will enable patients to receive targeted interventions and management plans.
• Personalized Care Plans: AI will assist primary care doctors and care teams in developing personalized care plans based on a patient’s risk profile. AI can also monitor patients’ progress and alert care teams to any concerning developments, leading to timely action.
• Patient Self-Management: AI tools can help patients self-manage their conditions by providing health status monitoring, timely reminders for medication, healthy eating, physical activity, and check-ups. This empowers patients to improve their health outcomes.
• Digitization and Standardization of Medical Data: CoE-AI will undertake the digitalization and standardization of medical data, facilitating precision medicine and research and development in healthcare.

4.1.3.3 Intelligent Learning and Assessment using AI

As per the Policy, To cater to individual learning needs and aptitudes, CoE-AI will develop a comprehensive program for personalized learning and assessment. The program will use AI-based adaptive skill assessment and training based on a modern curriculum in public-private partnerships. It will be initially launched as a pilot in public institutions in the Federal Capital and then expanded nationwide based on outcomes.

4.1.3.4 National IoT Data Cluster

CoE-AI will establish IoT cloud-based services at subsidized rates, supporting academic learning and offering state-of-the-art facilities for predictive analytics and insights generation in various industries. The initiative will invite local and international companies to provide IoT-based cloud services competitively.

4.1.3.5 Agriculture Supply Chain Optimization

CoE-AI will provide predictive analytics for optimizing the agricultural supply chain, reducing wastage, and improving inventory management. Additionally, weather prediction systems will assist farmers in making informed decisions about cultivation and harvest. IoTs and sensor technology will be employed for soil monitoring systems.

4.1.3.6 E-Khidmat Centers 2.0

CoE-AI will develop a smart citizen portal to streamline municipal service delivery, improve customer service quality, and automate paperwork. Chatbots will provide quick responses, and centralized service delivery will reduce employee burden and expenses.

4.1.3.7 Weather Prediction and Analysis using AI

CoE-AI will establish a weather prediction system based on machine learning models, dividing the country into regions with similar microclimates. This system will assist in preparedness against calamities, agricultural operations, and renewable energy solutions.

4.1.3.8 Intelligent Communications Systems (Road/Rail/Air)

AI will be utilized to predict commuters’ behavior and optimize transportation availability and revenue growth. It will assess affordability and accessibility of commuters and plan resources accordingly. AI will also help in assessing future demographics for transport planning, maintenance, and safety in road, rail, and air transport systems.

4.1.4 Data and Computational Infrastructure

This initiative focuses on standardizing data from State-Owned Enterprises (SoEs), boards, and civil authorities to apply AI-based algorithms for improved service provisioning and analytics. The AI Regulatory Directorate (ARD) will direct identified public sector entities and authorities to standardize their data efficiently and securely. CoE-AI will provide necessary infrastructure and partnerships with international AI platforms to process public data securely, and entities granted access to public data will be subject to scrutiny to ensure transparency and sustainability.

The initiative also emphasizes utilizing existing computational resources for data control and processing and encourages partnerships with local and international AI entities. Local startups, SMEs, and enterprises will be allowed to use the infrastructure and data in alignment with ARD directives. Security protocols will be defined to secure public data by pseudonymization before commissioning the AI market cloud on a national scale, creating a safe and secure data-sharing ecosystem. Policies will be formulated to maintain data standards based on industry and academia needs, with ARD responsible for data standards and processing requests.

4.2.1 Public Awareness of AI

To effectively adopt AI, the awareness of AI must be raised at the grassroots level to help individuals assess the necessity of data sharing and improve processing and exchange. A nationwide awareness program will be created to educate users about AI and its applications. The Ministry of IT & Telecom will devise a National Awareness Program for Personal AI adaptation, engaging various departments, PTA, and other institutions for maximum outreach through different media channels. The program will include technical information, breach scenarios, preventive measures, and more, distributed in various forms, including audible, visual, and readable content in national and regional languages. Special attention will be given to reaching the silver segment of society, marginalized women, and Persons with Disabilities (PWDs).

4.2.2 Research and Development

R&D in AI and allied technologies will be supported through grants for projects in academia and the private sector. The CoE-AI will enable at least 400 projects with fiscal and logistical support of up to PKR 1 million per project. Research and thesis support in AI will also be provided to promote potentially new concepts and models. The CoE-AI will offer fiscal, logistical, and data support for research, thesis, and publications. Additionally, the CoE-AI will facilitate the registration of new patents in AI and allied technologies to enhance Pakistan’s global footprint in research and development, aiming for at least 400 patent registrations annually.

4.2.3 Algorithms, Data Science & AI in Basic Education

The policy document recognizes the importance of imparting a fundamental understanding of algorithms, data science, and AI right from the grassroots level. To achieve this, CoE-AI will hire a local/international consultant with expertise in high-tech curriculum development to create a National Curriculum in Algorithms, Data Sciences, AI, and Allied Technologies for students from the sixth to twelfth standard. The consultant will develop the curriculum based on best practices within six months, along with an adoption strategy, and present the working draft to CoE-AI for further consultation.

CoE-AI, in collaboration with the Ministry of Federal Education and Professional Training, will organize a working committee comprising STEM education experts from the industry and abroad, local educational boards, and school associations to review the National High-Tech Curriculum Draft and propose recommendations within three months. Once finalized, the curriculum will be translated into local languages and made available to relevant institutions within six months of draft finalization.

4.2.4 Skill Development of Marginalized Women & PWDs

The policy document recognizes the importance of providing skill development opportunities for marginalized women and Persons with Disabilities (PWDs). As part of the National High-Impact Skill Development Program, CoE-AI will create a special offshoot to engage marginalized women and PWDs through tailored coursework and educational means. The program will be executed through institutions such as Bait-ul-Maal and Ehsaas Program, and will also be made available on the DigiSkills.pk platform for maximum inclusivity nationwide. Special quotas for marginalized women and PWDs will be ensured in each cohort. The program’s progress will be closely monitored and evaluated through periodic studies to ensure sustained access for all.

4.3 Building Progressive & Trusted Environment

The policy document outlines the establishment of an AI Regulatory Directorate (ARD) under the National Commission for Personal Data Protection (NCPDP) to address human rights, data privacy, and responsible use of AI-based technologies. The ARD will be responsible for regulating AI-based initiatives, defining regulatory functions concerning data trust, providing advisory services, and addressing AI policy issues. It will also work with other stakeholders to promote research, entrepreneurship, and innovative AI applications in line with national interests. The ARD will develop guidelines for data sharing, encourage local businesses to adopt AI solutions, and participate in international efforts for standardization in AI. The ARD will ensure that AI systems align with human dignity, rights, and cultural norms and will increase awareness of AI’s effects and risks on the socio-economic structure. It will also establish governance mechanisms to ensure fairness, data privacy, ethical values control, and algorithmic accountability in AI studies.

4.3.2 Generative AI

As per the Policy Generative AI has significant potential in various creative fields, including art, music, literature, and journalism. However, it also raises ethical and societal concerns due to its ability to create convincing fake content, which can lead to disinformation and privacy breaches. To address these challenges, the proposed AI Regulatory Directorate (ARD) will provide regulatory guidelines to ensure ethical use and prevent the spread of disinformation and fake news. The CoE-AI will focus on indigenous research and innovation in Generative AI, collaborating with international platforms like OpenAI. Additionally, it will work with the Higher Education Commission and the Ministry of Science and Technology to establish best practices for using Generative AI in academia.

4.3.3 Sandboxing for AI Deployment

Sandboxing provides a controlled environment for testing innovative technologies and allows developers and regulators to interact directly. The proposed intervention includes establishing a regulatory sandbox that will benefit at least 20 enterprises by 2025. This sandbox will enable real-world testing and discussions on ethical and legal scenarios related to AI applications. Additionally, at least 50 municipalities across the country will actively use AI-based technologies within smart city applications to encourage participation and adoption.

4.3.4 Supporting International Collaboration

International collaboration in AI-based research and innovative solutions will complement Pakistan’s efforts in AI proliferation. By collaborating with international experts and institutions, Pakistan can benefit from advancements and expertise in AI and upgrade its technological infrastructure to international standards. The proposed intervention is to launch at least five project calls in AI by 2025, prioritizing international collaboration. This will provide academia and industry in Pakistan with fresh perspectives and knowledge about the latest developments in AI from the global community.

4.4.1 Role of National IT Boards in Transformation

The National Information Technology Board (NITB) and provincial IT boards have a crucial role in data digitization and standardization across various sectors and industries. They will design roadmaps for transformations in different sectors based on their readiness for AI adoption, provide specialized training courses and certifications, and implement AI maturity models and project management guidelines in institutions and organizations to ensure compliance with international best practices.

4.4.2 Industrial Transformation

AI and allied technologies, including IoT, can significantly enhance industrial capacity building and efficiency. Introducing IoT in industries can streamline assembly lines, improve asset management, enhance machine efficiency, optimize inventory management, and promote sustainable energy usage. The Federation of Pakistan Chamber of Commerce and Industry (FPCCI) will play a role in promoting AI and IoT adoption in industries through awareness campaigns, seminars, workshops, and training courses.

4.4.3 Public/Private Sector Evolution

The CoE-AI and ARD will work collaboratively to ensure a smooth AI adoption process in the public and private sectors. The National Artificial Intelligence Fund (NAIF) will fuel research and innovative endeavors, and collaborative investments from the private sector and international bodies will support the commercialization of AI products within Pakistan’s institutional and industrial landscape.

5 Policy Implementation & Review

As per the Policy, to ensure effective implementation and continuous improvement of the AI policy, a robust governance structure is proposed. This structure includes a Steering/Management Committee, Working Groups, and a Policy Implementation Cell. The roles and responsibilities of each component are defined as follows:

5.1 Steering/Management Committee

• Comprising representatives from the government, academia, industry, and civil society with equal representation.
• Chaired by the Secretary of IT & Telecom or the acting chair in the absence of the Secretary.
• Analyzes and directs the progress and performance of different working groups, the policy implementation cell, and organizations resulting from policy interventions.
• Provides guidelines for new and existing initiatives to the National AI Fund in alignment with the policy.

5.2 Working Groups

• Four working groups representing the four policy drivers, with participation from all stakeholders.
• Collaborates with international organizations and academia to update the AI Policy based on global best practices.
• Provides recommendations on adopting international regulation standards to the steering committee.
• Assists in research areas where policy measures are needed, identifies obstacles, and proposes solutions.
• Monitors and evaluates their respective domains, communicating recommendations to the steering committee.
• Liaises with institutions like NAIF and CoE-AI to monitor and evaluate the operations of various organizations.

5.3 Policy Implementation Cell

• Oversees day-to-day operations to implement policy guidelines.
• Provides operational support to the working groups and management committee for smooth implementation.
• Works closely with the working groups and steering committee to monitor progress and provide recommendations on a monthly basis.
• Operates under the purview of MoITT and reports to Member IT.

5.4 Review Procedure & Timeline

• Comprehensive policy review every three years by the Ministry of IT & Telecom.
• Progress/implementation review of policy directives every six months by the Steering Committee.
• Working groups meet every three months to suggest recommendations to the steering committee.
• The Policy Implementation Cell starts operating immediately upon policy promulgation.
• Biannual meetings of the Steering/Management Committee to assess progress and tasks reported by working groups and Policy Implementation Cell.
• Appropriate stipends provided to the working groups and steering/management committee for effective functioning.
• Monitoring and evaluation may involve third-party organizations contracted on an annual basis.

6 Definitions

• Triple Helix Partnership: A model involving collaboration between universities, governments, and industries to promote knowledge sharing and economic development.
• Small and Medium-Sized Enterprises (SME): Businesses and organizations between small office home office and larger enterprises.
• Sustainable Development Goals (SDG): A set of 17 global goals designed to achieve a better and sustainable future for all.
• Common Operating Environment (COE): A standard data collection and processing mechanism within an IT architecture to promote interoperability.
• Key Performance Indicator (KPI): A quantifiable measure of performance for a specific objective.
• Startups: Innovative, high-growth potential, technology-based businesses that require unique skills and resources for commercialization and sustainable growth.

Annex I of the Policy – Sectoral Survey Regarding AI Adoption, Challenges, and Opportunities in Pakistan

8.1 Sectoral Identification

To strategize AI proliferation in Pakistan, it is essential to identify sectors that can benefit the most from AI adoption and progress. The sectors are chosen based on criteria that indicate their potential for AI adoption and the positive impact it can have on their operations. The identified societal areas are as follows:

• Agriculture: AI can revolutionize the agricultural sector through process automation, traceability, and data analysis from satellites and weather predictions.
• Healthcare: AI can benefit healthcare through diagnostics, outbreak prevention, R&D in pharmaceuticals, personalized healthcare, and precision medicine.
• Manufacturing: AI and robotics can enhance efficiency and safety in the manufacturing industry.
• Retail & Commerce: AI can streamline inventory and supply chain management in the retail and commerce sector.
• IT Industry: The IT industry can leverage AI for re-engineering processes, predictive analytics, and growth opportunities in global markets.
• Financial Sector: AI can improve customer experience, identify investment opportunities, and enhance data safety in the financial sector.
• Academia: AI can assist in teaching and skills development in educational institutions.
• Government Sector: AI can improve efficiency and decision-making in public services.
• Defense Sector: AI can bring revolutionary improvements in various defense sub-sectors, enhancing national defense capabilities.

Survey Statistics as per the Policy

The Ministry of IT & Telecom conducted a sectoral survey to assess the state of AI in Pakistan. The participating organizations’ contributions are summarized, and key findings in each sector are highlighted.

Note: The specific statistics and findings from the sectoral survey are not provided in the provided text. They would likely include data on AI adoption rates, challenges faced in implementing AI, opportunities for growth and innovation, and the overall readiness of each sector to embrace AI technologies.

Table 2: Society Challenges Reflected by the Survey Results

Note: The above table adopted from the Policy summarizes the evidence gathered from the sectoral survey regarding various challenges and opportunities in the adoption of AI in Pakistan. It includes statistics related to AI readiness, education, research and development, investment, AI governance/regulations, and other market challenges faced in implementing AI technologies. The challenges include awareness and adaptation, data standardization and accessibility, computational infrastructure needs, and ethical considerations for AI adoption.

As per the policy, the market challenges related to AI adoption in Pakistan can be categorized into four groups: Awareness and Adaptation Challenges (G1), Data Standardization and Accessibility (G2), Computational Infrastructure Situation and Needs (G3), and Ethical Challenges (G4). These challenges are identified based on the evidence gathered from the sectoral survey. Let’s take a closer look at each group:

G1 – Awareness and Adaptation Challenges:

• Lack of Awareness: There is a need to raise awareness about the potential of AI in solving routine problems and its applications in various sectors. An awareness campaign is necessary to promote data control and sharing, basic skills education, and high-impact skilling at all levels of the community.
• Access to Skilled Human Resources: To enable AI and allied technologies, skilled human resources with a broader understanding of AI applications and best practices are required. The access to data for training these skilled resources is essential for their effective use of AI-based models.
• Workforce Adaptation: The adoption of AI can disrupt traditional work processes in public and private sectors, potentially leading to job losses for individuals not skilled in AI-based technologies. Customized learning programs, workshops, and courses are needed to help individuals adapt to the digital transformation in workplaces.
• Curriculum Development: There is a need to include short courses, boot camps, and diplomas focusing on AI-based curricula at various education levels, including bachelor’s, master’s, and Ph.D. degrees.

G2 – Data Standardization and Accessibility:

• Data Standardization: The massive datasets available in various sectors need to be digitized and organized into a standardized format to make them relevant for data analytics and other data processing mechanisms. Currently, sectors like manufacturing, agriculture, retail & commerce, and healthcare lag in data standardization.
• Data Accessibility: Accessibility to data is a significant hurdle in AI adoption, as many datasets are not available or accessible in appropriate formats. The lack of accessibility hinders AI-based research and development and predictive analytics.
• Data Collection Practices: Poor data collection practices are reported by stakeholders, affecting successful AI adoption.

G3 – Computational Infrastructure Situation and Needs:

• Limited Computational Resources: Availability of computational equipment and facilities is necessary for training human capital and modernizing processes for AI adoption. Currently, there is a shared obstacle of limited computational resources in various sectors of Pakistan.
• Optimal Use of Computational Resources: The computational resources available are not always utilized optimally, leading to the need for a national/provincial data repository for public sector data. Training on high-performance computing technologies is essential for meeting local computational requirements.

G4 – Ethical Challenges:

• Safe and Secure Data Sharing: Ensuring safe and secure data sharing is a primary concern in AI adoption. Establishing an effective regulatory system is crucial for ethical use of AI and related technologies and to build trust among users.
• Risks Associated with AI Adoption: Stakeholders acknowledge the medium to high risks associated with AI adoption, including malicious intent, data pollution, data model theft, fairness, precision deterioration, and invasion of privacy.

Addressing these market challenges and implementing necessary interventions will be essential for promoting the successful adoption and integration of AI technologies in Pakistan’s various sectors.

A critical analysis

The draft policy on Artificial Intelligence in Pakistan suffers from several critical flaws that hinder its effectiveness and potential impact on the country’s AI landscape:

1. Lack of National Outlook and Inclusivity: The policy should have been developed through an inclusive, multi-stakeholder dialogue involving industry experts and academia. However, it appears that the draft was prepared with limited coordination, possibly by handpicked consultants, leading to a lack of diverse perspectives and a cohesive approach towards AI.
2. Bureaucratic Aversion to Digitisation: There seems to be a resistance within the bureaucratic system to fully embrace digitization and AI adoption. This mindset needs to be overcome through strong political supervision and a commitment to phasing out outdated practices, particularly in the “clerical corps,” to pave the way for effective AI implementation.
3. Internet Access and Connectivity Restrictions: The policy should promote open and unrestricted Internet access, essential for fostering innovation and AI development. Instead, the state is urged to find targeted approaches for managing web links rather than resorting to blanket shutdowns, which can stifle progress.
4. Pointless Goal-Setting: The policy sets fixed deadlines and cut-off dates for AI adoption, which are impractical and counterproductive. AI is a rapidly evolving field, and setting rigid timelines can lead to compromised quality and improper means to achieve goals.
5. Lack of Distributed Efforts in AI R&D: The policy lacks a holistic and integrated approach to AI research and development. Instead of creating a new, separate entity, an independent National Coordinator should be appointed to coordinate efforts among existing channels, ensuring better synergy and resource utilization.

Moreover, the policy’s implementation and projects lack transparency and evaluation, particularly in academia-centered initiatives. The discussion around AI appears to be more rhetorical than substantive, indicating a lack of concrete progress in practical terms.

The absence of inclusive multi-stakeholder discussions from the federal government and a focus on securitisation-oriented funding for AI-related R&D projects raise concerns about the dominance of certain institutions and the limited national orientation towards AI adoption.Furthermore, the delay in addressing the Personal Data Protection Act (PDPA) draft and the lack of seriousness accorded to this critical legislation indicate a disregard for data security and privacy, fundamental for a robust AI ecosystem.

Overall, the draft policy appears to be a hasty  attempt, lacking in practical insights and a comprehensive SWOT analysis of existing initiatives. Informed policy-making demands a thorough understanding of strengths and weaknesses, which seems to be missing from this document.Without genuine efforts to involve academic experts and industry veterans, the policy may struggle to achieve its intended objectives, leaving politicians seemingly disengaged and oblivious to the real potential of AI for the nation.

************We will keep this page updated with updates on the Policy and further legal developments************

For any concerns or questions please reach us at [email protected]