The newly signed memorandum of understanding between the Indus Institute for AI & Digital Policy (IIADP) and Turkish health-policy expert Dr Kemal Aydin is more than a conventional academic partnership: it is an attempt to give Pakistan, Türkiye and the wider D-8 community a more coordinated voice in the rapidly evolving world of artificial intelligence policy, digital governance and public-sector innovation.
The agreement creates a long-term framework for joint research, policy consultations, international conferences, institutional networking and capacity-building programmes. Its stated ambition is to help D-8 countries move beyond isolated technology projects and toward evidence-based approaches to AI adoption, data governance and sustainable digital development.
For a region where AI discussions are often dominated by infrastructure gaps, skills shortages and imported technology platforms, the MoU places policy at the center of the conversation. That matters. The greatest challenge is no longer simply whether governments can access AI tools, but whether they can deploy them responsibly, securely and in ways that deliver measurable public value.
The MoU links IIADP with Dr Aydin’s policy experience in health and public administration, creating a platform that spans research, governance and institutional engagement. While the agreement does not itself create a new D-8 authority or binding regulatory body, it establishes a practical channel for people and institutions working on AI governance to exchange knowledge across borders.
That distinction is important. Memorandums of understanding are generally framework agreements, not detailed implementation plans. Their value depends on what follows: shared research outputs, recurring dialogue, training programmes, government participation and specific projects with accountable timelines.
Still, the structure outlined in this agreement is well suited to the kind of cross-border coordination that AI policy increasingly requires. Artificial intelligence systems do not stay neatly within national boundaries. Data may be collected in one country, processed in another, hosted on infrastructure owned elsewhere and used to make decisions affecting citizens in a fourth jurisdiction.
A regional approach can therefore help participating countries address common questions:
AI and digital governance add a new layer to that agenda.
The D-8 countries do not all have identical digital ecosystems, regulatory structures or technology capabilities. Some have stronger startup communities, larger telecommunications markets, more mature government digital services or deeper pools of technical talent. Others face more acute connectivity, affordability and institutional-capacity challenges.
That variation can be an advantage if cooperation is designed well. Instead of every country attempting to independently solve the same policy problems, a D-8 AI policy network could identify successful models, adapt them locally and avoid repeating expensive mistakes.
A government can launch an AI pilot relatively quickly. It is much harder to ensure that the system is accurate, fair, explainable, secure, accessible and legally defensible over time.
A public-sector AI tool may appear efficient while still producing harmful outcomes if its training data is incomplete, biased or outdated. A health triage system might disadvantage people from underrepresented communities. An automated benefits system could make opaque eligibility decisions. A predictive model used in policing or border management could amplify historical inequalities.
This is why AI governance must become a core public-sector capability rather than an afterthought attached to technology procurement.
The MoU’s focus on policy research and strategic dialogue suggests an awareness that responsible AI deployment requires more than enthusiasm for automation. It requires institutional rules, technical review processes and a willingness to test systems before they affect the public.
The IIADP partnership fits into this wider direction by emphasizing international policy engagement. National strategies are necessary, but AI governance cannot be built entirely within domestic borders. Many core issues are international by nature, including cloud infrastructure, cross-border data transfers, language-model development, semiconductor supply chains, cybersecurity, platform regulation and global standards.
For Pakistan, closer collaboration with Türkiye and other D-8 partners offers several potential benefits.
A cross-border network can expand the available pool of knowledge. It can also provide a forum for comparing how different countries handle similar challenges, from protecting health data to building digital public infrastructure.
A coordinated D-8 position could help ensure that the priorities of emerging economies receive more attention. These priorities may include:
The MoU could support practical work in areas such as algorithmic impact assessments, data-sharing agreements, procurement standards, incident reporting and model-evaluation procedures. These may sound procedural, but they are the mechanisms that determine whether responsible AI principles become operational.
AI can support medical imaging, disease surveillance, hospital capacity planning, patient triage, clinical documentation and population-health analysis. It can help health systems identify patterns that would be difficult to detect manually, especially where staff are stretched and patient volumes are high.
But healthcare AI is not just another software category.
Governments and healthcare providers must also examine whether a system works reliably across different populations. A model trained primarily on data from wealthier countries or narrow demographic groups may deliver less accurate results for patients elsewhere.
The consequences of poor performance can be severe. A flawed entertainment recommendation is inconvenient. A flawed medical recommendation can affect diagnosis, treatment and patient safety.
That makes healthcare an ideal proving ground for a more mature approach to AI governance. It requires multidisciplinary oversight, clinical validation, clear accountability and meaningful human review.
Without reliable, lawful and well-governed data, public-sector AI projects are unlikely to produce dependable outcomes. Poor-quality data can produce poor-quality models. Incomplete data can exclude communities. Insecure data can expose citizens to harm.
The MoU’s inclusion of data policy is therefore one of its most consequential elements.
Effective data governance should address several fundamentals:
The stronger approach is to create adaptable frameworks that establish common baseline protections while allowing countries to tailor implementation.
At its best, public-sector innovation means redesigning services around citizens’ needs, reducing administrative friction and using technology to improve access, speed and quality. At its worst, it can become a label used to justify untested automation or expensive pilot projects that never scale.
AI should not be adopted simply because it is fashionable. Public institutions should begin with a clearly defined problem.
Potential examples include:
This is where policy research and consultation can add real value. A regional network can help governments develop a shared vocabulary for classifying risk and deciding when an AI application requires independent review.
A successful implementation agenda would produce visible, reusable work that public institutions, researchers and policymakers can rely on. It would also create opportunities for younger researchers and civil servants to develop expertise in an area that is becoming central to modern governance.
Better indicators would include:
Governments need the ability to inspect systems, investigate incidents, demand corrective action and suspend high-risk tools when necessary. Without those powers, responsible AI commitments may remain aspirational.
External partnerships are not inherently negative. They can accelerate access to useful technology. However, public institutions should avoid contracts that leave them unable to understand, audit, migrate or maintain critical AI systems.
Interoperability, open standards, data portability and transparent procurement can reduce this risk.
A credible governance model should include public-interest voices, disability advocates, consumer groups, researchers, journalists and communities affected by automated decisions. Consultation should not be a formality held after a system has already been selected.
The answer is not to lower standards. It is to build phased, realistic pathways that help institutions move from basic AI awareness to structured risk management and eventually to more sophisticated oversight.
Pakistan brings a large youthful population, a growing technology sector, an expanding digital-policy environment and a strong need for AI solutions that work across diverse languages and public-service conditions. Türkiye brings substantial experience in public administration, healthcare policy, technology development and regional engagement.
Together, they can help advance a model of cooperation that treats AI not only as an engine of economic growth, but also as a public-policy responsibility.
That framing is increasingly necessary. AI will influence how states deliver services, regulate markets, protect data, train workers and respond to emergencies. Countries that invest only in adoption may find themselves dependent on systems they cannot meaningfully govern. Countries that invest in governance without building practical capability may miss economic and social opportunities.
The balance is difficult, but it is achievable: encourage innovation, build domestic capacity, protect citizens and preserve accountable human decision-making.
The agreement’s emphasis on research, dialogue, networking and capacity building reflects a realistic understanding of the challenge. No country can create a mature AI governance system overnight, and no single institution can address the issue alone.
Its real value will depend on whether the partnership produces durable mechanisms for cooperation: useful research, skilled officials, credible policy tools and public-sector projects that demonstrate responsible AI in practice. If it does, the MoU could help turn the D-8 community into a more active participant in shaping the rules of the AI era rather than merely responding to decisions made elsewhere.
For Pakistan, Türkiye and their regional partners, that is the strategic opportunity behind this agreement: not simply to discuss artificial intelligence, but to build the governance capacity needed to ensure that AI serves development, public trust and long-term national interests.
The agreement creates a long-term framework for joint research, policy consultations, international conferences, institutional networking and capacity-building programmes. Its stated ambition is to help D-8 countries move beyond isolated technology projects and toward evidence-based approaches to AI adoption, data governance and sustainable digital development.
For a region where AI discussions are often dominated by infrastructure gaps, skills shortages and imported technology platforms, the MoU places policy at the center of the conversation. That matters. The greatest challenge is no longer simply whether governments can access AI tools, but whether they can deploy them responsibly, securely and in ways that deliver measurable public value.
A Policy Partnership With Regional Ambition
The MoU links IIADP with Dr Aydin’s policy experience in health and public administration, creating a platform that spans research, governance and institutional engagement. While the agreement does not itself create a new D-8 authority or binding regulatory body, it establishes a practical channel for people and institutions working on AI governance to exchange knowledge across borders.That distinction is important. Memorandums of understanding are generally framework agreements, not detailed implementation plans. Their value depends on what follows: shared research outputs, recurring dialogue, training programmes, government participation and specific projects with accountable timelines.
Still, the structure outlined in this agreement is well suited to the kind of cross-border coordination that AI policy increasingly requires. Artificial intelligence systems do not stay neatly within national boundaries. Data may be collected in one country, processed in another, hosted on infrastructure owned elsewhere and used to make decisions affecting citizens in a fourth jurisdiction.
A regional approach can therefore help participating countries address common questions:
- How should public institutions evaluate AI systems before deployment?
- What data protections should apply when citizens’ information is used to train or operate AI tools?
- How can governments procure AI services without becoming locked into a single vendor?
- What safeguards are necessary when algorithms influence healthcare, welfare, education, policing or employment?
- How can countries develop local-language, culturally relevant and economically useful AI systems?
- What skills do civil servants need to oversee AI responsibly?
Why the D-8 Context Matters
The D-8 Organization for Economic Cooperation was created to strengthen economic collaboration among developing countries with large populations, diverse markets and significant development ambitions. Its traditional focus has included trade, agriculture, industry, transport, tourism and small and medium-sized enterprises.AI and digital governance add a new layer to that agenda.
The D-8 countries do not all have identical digital ecosystems, regulatory structures or technology capabilities. Some have stronger startup communities, larger telecommunications markets, more mature government digital services or deeper pools of technical talent. Others face more acute connectivity, affordability and institutional-capacity challenges.
That variation can be an advantage if cooperation is designed well. Instead of every country attempting to independently solve the same policy problems, a D-8 AI policy network could identify successful models, adapt them locally and avoid repeating expensive mistakes.
From Technology Adoption to Governance Capacity
For years, digital transformation was frequently measured through visible projects: online portals, mobile apps, digital payments, cloud migration and national broadband targets. Those remain important, but AI introduces harder questions.A government can launch an AI pilot relatively quickly. It is much harder to ensure that the system is accurate, fair, explainable, secure, accessible and legally defensible over time.
A public-sector AI tool may appear efficient while still producing harmful outcomes if its training data is incomplete, biased or outdated. A health triage system might disadvantage people from underrepresented communities. An automated benefits system could make opaque eligibility decisions. A predictive model used in policing or border management could amplify historical inequalities.
This is why AI governance must become a core public-sector capability rather than an afterthought attached to technology procurement.
The MoU’s focus on policy research and strategic dialogue suggests an awareness that responsible AI deployment requires more than enthusiasm for automation. It requires institutional rules, technical review processes and a willingness to test systems before they affect the public.
Pakistan’s Expanding AI Policy Landscape
Pakistan has been moving toward a more structured national AI agenda, including the approval of its National Artificial Intelligence Policy 2025 and subsequent efforts to translate policy objectives into institutional capacity. These initiatives aim to support AI research, workforce development, public-sector adoption, innovation and responsible use.The IIADP partnership fits into this wider direction by emphasizing international policy engagement. National strategies are necessary, but AI governance cannot be built entirely within domestic borders. Many core issues are international by nature, including cloud infrastructure, cross-border data transfers, language-model development, semiconductor supply chains, cybersecurity, platform regulation and global standards.
For Pakistan, closer collaboration with Türkiye and other D-8 partners offers several potential benefits.
A Broader Pool of Expertise
No single government has all of the technical, legal and administrative expertise needed to govern AI effectively. Policymakers need input from computer scientists, data-protection specialists, economists, ethicists, educators, cybersecurity professionals, civil-society groups and sector experts.A cross-border network can expand the available pool of knowledge. It can also provide a forum for comparing how different countries handle similar challenges, from protecting health data to building digital public infrastructure.
Greater Influence in Global AI Discussions
Much of the international debate around AI governance is shaped by major technology powers, large platform companies and institutions with substantial technical resources. Developing countries frequently face the prospect of adopting standards created elsewhere, even when those standards do not fully reflect local languages, economic realities or public-service needs.A coordinated D-8 position could help ensure that the priorities of emerging economies receive more attention. These priorities may include:
- Affordable access to computing infrastructure
- Responsible technology transfer
- Data sovereignty and local control
- Multilingual AI tools
- Digital inclusion for rural and underserved populations
- Skills development for young people and civil servants
- AI applications for health, agriculture, education and climate resilience
- Fair access to global AI markets
A Route to Practical Policy Learning
The most useful AI policy discussions are connected to real public-sector problems. Governments need guidance that helps them decide whether an AI system should be deployed, how it should be tested and who remains accountable when it fails.The MoU could support practical work in areas such as algorithmic impact assessments, data-sharing agreements, procurement standards, incident reporting and model-evaluation procedures. These may sound procedural, but they are the mechanisms that determine whether responsible AI principles become operational.
The Health Policy Dimension Could Be Especially Significant
Dr Aydin’s background as a Turkish health-policy expert and former Ministry of Health official gives the agreement an important sectoral perspective. Healthcare is one of the areas where AI offers substantial potential while also carrying unusually high risks.AI can support medical imaging, disease surveillance, hospital capacity planning, patient triage, clinical documentation and population-health analysis. It can help health systems identify patterns that would be difficult to detect manually, especially where staff are stretched and patient volumes are high.
But healthcare AI is not just another software category.
Sensitive Data, Serious Consequences
Health information is among the most sensitive categories of personal data. Any AI system that processes medical records, diagnostic images, genetic information or patient communications requires strong protections against unauthorized access, misuse and re-identification.Governments and healthcare providers must also examine whether a system works reliably across different populations. A model trained primarily on data from wealthier countries or narrow demographic groups may deliver less accurate results for patients elsewhere.
The consequences of poor performance can be severe. A flawed entertainment recommendation is inconvenient. A flawed medical recommendation can affect diagnosis, treatment and patient safety.
That makes healthcare an ideal proving ground for a more mature approach to AI governance. It requires multidisciplinary oversight, clinical validation, clear accountability and meaningful human review.
Opportunities for D-8 Cooperation in Digital Health
A D-8 policy network could explore common principles for AI-enabled health systems without forcing countries into a single technical model. Potential areas of work include:- Clinical validation standards for AI-assisted diagnostic tools.
- Health-data governance frameworks that balance research benefits with patient privacy.
- Interoperability principles for exchanging health information securely.
- Guidelines for human oversight in clinical and administrative decision-making.
- Capacity-building programmes for health officials, regulators and hospital administrators.
- Evaluation methods for detecting bias, unsafe recommendations and performance gaps.
- Procurement safeguards for hospitals purchasing AI services from commercial suppliers.
Data Policy Is the Foundation Beneath AI Policy
Artificial intelligence depends on data. That simple fact is often obscured by the excitement surrounding chatbots, image generators and autonomous systems.Without reliable, lawful and well-governed data, public-sector AI projects are unlikely to produce dependable outcomes. Poor-quality data can produce poor-quality models. Incomplete data can exclude communities. Insecure data can expose citizens to harm.
The MoU’s inclusion of data policy is therefore one of its most consequential elements.
The Need for Clear Data Governance
Data governance refers to the rules, roles, processes and technical protections that determine how information is collected, stored, accessed, shared, retained and deleted. In an AI context, it also involves knowing where data comes from, whether it can legally be used and how it affects model behavior.Effective data governance should address several fundamentals:
- Purpose limitation: Data should be used for clearly defined and legitimate purposes.
- Data minimization: Institutions should avoid collecting more personal information than necessary.
- Quality controls: Data should be accurate, current and suitable for the intended use.
- Security protections: Sensitive information must be protected against breaches and misuse.
- Access controls: Only authorized users should be able to view or process data.
- Transparency: Citizens should understand, where appropriate, how their data is being used.
- Retention rules: Data should not be retained indefinitely without a justified reason.
- Accountability: Institutions need identifiable officials responsible for compliance.
The stronger approach is to create adaptable frameworks that establish common baseline protections while allowing countries to tailor implementation.
The Importance of Public-Sector Innovation
The agreement also highlights public-sector innovation, a term that can mean very different things depending on how it is applied.At its best, public-sector innovation means redesigning services around citizens’ needs, reducing administrative friction and using technology to improve access, speed and quality. At its worst, it can become a label used to justify untested automation or expensive pilot projects that never scale.
AI should not be adopted simply because it is fashionable. Public institutions should begin with a clearly defined problem.
Start With the Problem, Not the Model
A useful AI project should answer a practical question: what public-service challenge is being addressed, and why is AI the right tool?Potential examples include:
- Identifying crop disease risks from field imagery
- Improving multilingual access to government information
- Forecasting demand for hospital supplies
- Detecting anomalies in public procurement data
- Supporting teachers with personalized learning resources
- Reducing document-processing backlogs
- Improving disaster-response coordination
- Analyzing infrastructure maintenance needs
This is where policy research and consultation can add real value. A regional network can help governments develop a shared vocabulary for classifying risk and deciding when an AI application requires independent review.
What Success Would Look Like
The signing of the MoU is a meaningful starting point, but its long-term significance will be measured by outcomes rather than announcements.A successful implementation agenda would produce visible, reusable work that public institutions, researchers and policymakers can rely on. It would also create opportunities for younger researchers and civil servants to develop expertise in an area that is becoming central to modern governance.
Practical Deliverables Worth Pursuing
The partnership could achieve tangible impact by focusing on a manageable set of deliverables.- A D-8 AI governance policy brief series addressing shared challenges in data protection, public procurement, algorithmic accountability and digital inclusion.
- A regional AI policy observatory tracking laws, strategies, public-sector projects and emerging regulatory approaches across participating countries.
- Training modules for civil servants on AI literacy, risk assessment, procurement and oversight.
- Model procurement clauses that require vendors to provide information about data use, security, model limitations and audit access.
- Algorithmic impact assessment templates for high-risk government AI projects.
- Annual policy dialogues or conferences connecting researchers, regulators, technologists and public administrators.
- Sector-specific working groups focused on health, agriculture, education, digital identity and public finance.
- Fellowship and exchange programmes for researchers and early-career policy professionals.
Metrics Matter
The partnership should also resist the tendency to measure success only by the number of workshops held or memorandums signed. Those activities can be valuable, but they are inputs, not outcomes.Better indicators would include:
- Number of policy recommendations adopted by public institutions
- Number of civil servants trained in AI governance
- Number of cross-border research projects completed
- Publication of usable governance tools and model frameworks
- Participation from universities, civil society and industry
- Evidence that public-sector AI projects are being evaluated for risk and impact
- Improvements in access, quality or efficiency for specific public services
Risks That Cannot Be Ignored
The growing momentum around AI policy also brings risks. Regional cooperation can amplify good practice, but it can also reproduce weaknesses if the work becomes overly focused on statements of intent rather than implementation.The Risk of Policy Without Enforcement
AI principles are easy to endorse. Terms such as fairness, transparency, privacy and accountability appear in almost every policy document. The difficult part is creating enforcement mechanisms when systems are deployed.Governments need the ability to inspect systems, investigate incidents, demand corrective action and suspend high-risk tools when necessary. Without those powers, responsible AI commitments may remain aspirational.
The Risk of Vendor Dependence
Many governments lack the computing infrastructure and technical capacity to develop advanced AI systems independently. That can lead to dependence on a small number of foreign cloud providers, model developers or consulting firms.External partnerships are not inherently negative. They can accelerate access to useful technology. However, public institutions should avoid contracts that leave them unable to understand, audit, migrate or maintain critical AI systems.
Interoperability, open standards, data portability and transparent procurement can reduce this risk.
The Risk of Excluding Citizens
AI policy is often discussed by governments, companies and technical experts. Yet citizens are the people most affected when AI is used in public services.A credible governance model should include public-interest voices, disability advocates, consumer groups, researchers, journalists and communities affected by automated decisions. Consultation should not be a formality held after a system has already been selected.
The Risk of Uneven Capacity
Not every D-8 country will be ready to implement the same standards at the same pace. Differences in digital infrastructure, legal frameworks, public funding and technical expertise are real.The answer is not to lower standards. It is to build phased, realistic pathways that help institutions move from basic AI awareness to structured risk management and eventually to more sophisticated oversight.
A Strategic Opening for Pakistan and Türkiye
Pakistan and Türkiye have an opportunity to use this MoU as a bridge between national digital strategies and a broader regional AI governance agenda.Pakistan brings a large youthful population, a growing technology sector, an expanding digital-policy environment and a strong need for AI solutions that work across diverse languages and public-service conditions. Türkiye brings substantial experience in public administration, healthcare policy, technology development and regional engagement.
Together, they can help advance a model of cooperation that treats AI not only as an engine of economic growth, but also as a public-policy responsibility.
That framing is increasingly necessary. AI will influence how states deliver services, regulate markets, protect data, train workers and respond to emergencies. Countries that invest only in adoption may find themselves dependent on systems they cannot meaningfully govern. Countries that invest in governance without building practical capability may miss economic and social opportunities.
The balance is difficult, but it is achievable: encourage innovation, build domestic capacity, protect citizens and preserve accountable human decision-making.
The Real Test Begins After the Signature
The IIADP and Dr Kemal Aydin MoU arrives at a moment when AI policy is shifting from abstract debate to institutional necessity. Governments are being asked to make consequential choices about data, public procurement, automated decision-making and digital sovereignty—often before laws, standards and technical capacity have fully caught up.The agreement’s emphasis on research, dialogue, networking and capacity building reflects a realistic understanding of the challenge. No country can create a mature AI governance system overnight, and no single institution can address the issue alone.
Its real value will depend on whether the partnership produces durable mechanisms for cooperation: useful research, skilled officials, credible policy tools and public-sector projects that demonstrate responsible AI in practice. If it does, the MoU could help turn the D-8 community into a more active participant in shaping the rules of the AI era rather than merely responding to decisions made elsewhere.
For Pakistan, Türkiye and their regional partners, that is the strategic opportunity behind this agreement: not simply to discuss artificial intelligence, but to build the governance capacity needed to ensure that AI serves development, public trust and long-term national interests.
References
- Primary source: The News Pakistan
Published: 2026-07-25T19:00:00+00:00
MoU on cooperation in AI policy signed
KARACHI: The Indus Institute for AI & Digital Policy and Dr Kemal Aydin, a Turkish health-policy expert and former Ministry of Health official, have signed a memorandum of understanding to...www.thenews.pk
