Pakistan brings in foreign AI experts as AI Council formation drags

Pakistan will bring foreign AI experts into its new council to jumpstart a stalled policy rollout. The real test: fixing HPC gaps and moving from plans to pilots, fast.

Categorized in: AI News Government
Published on: Feb 08, 2026
Pakistan brings in foreign AI experts as AI Council formation drags

Pakistan to add foreign AI experts to national council as policy rollout stalls

Pakistan will include international AI experts in its forthcoming AI Council to execute the National Artificial Intelligence Policy. The move is meant to avoid a purely bureaucratic body and bring practical expertise to the table, which officials say is a key reason the council's notification has been delayed.

The federal cabinet approved the National AI Policy in July 2025. Provinces were asked for input about a month ago, but the council-central to steering, coordinating, and regulating AI-has not yet been notified.

What the council will do

The proposed council will align AI development with national priorities, regulate emerging technologies, and coordinate across federal and provincial stakeholders. It is expected to include representatives from key ministries and regulators-Science and Technology, Foreign Affairs, Federal Education, the Higher Education Commission, and the Pakistan Telecommunication Authority-augmented by local and international AI experts.

The bottleneck: compute and infrastructure

Pakistan's AI ambitions are constrained by limited high-performance computing (HPC). Only three universities are reported to have functional HPC data centres. Telecom operators do run data centres, but most are built for conventional IT and network workloads, not training or serving advanced AI models.

Private investors are interested in AI-focused data centres, yet unclear policy and regulatory guardrails are slowing decisions. Without timely coordination and infrastructure planning, Pakistan risks lagging further behind in AI-undermining broader digital and economic goals. For context on global policy approaches, see the OECD AI Policy Observatory here.

What government teams can do now

  • Notify the council with clear criteria: Define expert selection, conflict-of-interest rules, and term limits to keep it focused and credible.
  • Stand up a small secretariat: Create working groups for compute and data, safety and regulation, skills and procurement, and provincial coordination.
  • Map national compute capacity: Inventory public and private data centres, quantify GPU/accelerator availability, and scope shared HPC services for universities and agencies.
  • Set interim rules for AI pilots: Issue short guidance on data privacy, security, model evaluation, and procurement so departments can start safe pilots.
  • Lock a consultation calendar: Publish timelines with provinces and regulators; commit to regular council meetings and public minutes.
  • Enable private investment: Outline fast-track approvals, grid access/energy incentives, and tax options for AI-grade data centres.
  • Prioritize quick wins in government: Document processing, citizen service assistants, and fraud/anomaly detection-run in sandboxed environments with audit trails.

90-day implementation checkpoints

  • Days 0-30: Issue council notification; appoint chair and secretariat; publish selection criteria; release interim AI pilot guidance.
  • Days 31-60: Complete compute inventory; identify at least two shared HPC sites; launch three inter-agency pilot projects with clear success metrics.
  • Days 61-90: Publish a regulatory workplan (privacy, safety, procurement templates); sign MOUs with provinces; announce an investment brief for AI-focused data centres.

Guardrails that matter

  • Expert balance: Combine domain experts (health, education, finance) with ML engineering and policy expertise; avoid tokenism.
  • Transparency: Publish membership, meeting notes, and conflict disclosures.
  • Data governance: Clarify data-sharing protocols, localization requirements, and cross-border access for research and services.
  • Security by default: Adopt secure data environments, access controls, and logging for all government AI pilots.
  • Procurement templates: Standardize clauses for model performance, bias testing, privacy, uptime, and exit options.

If your team needs practical upskilling to run safe, effective AI pilots in the public sector, see curated AI courses by job.

The bottom line

Bringing global experts into the council is the right call-if it accelerates real decisions. Pair the notification with immediate guidance, a compute plan, and visible pilot wins. Momentum now will decide whether the policy becomes a headline or delivers results across government.


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)