Summary

In 2024, Pakistan’s Ministry of Education partnered with Taleemabad through the NIETE programme to support teachers with AI-generated lesson plans, reaching 4,000 teachers and 98,000 students. EdTech Hub’s evaluation revealed that while AI improved consistency and saved time, flexibility and teacher agency remain crucial. The story highlights how designing with teachers—not just for them—can make AI a meaningful ally in improving teaching and learning experiences.

Key Data Points

The NIETE programme rolled out across 341 public schools, reaching 4,044 teachers and nearly 98,000 students.

The Programme at a Glance

In 2024, the Ministry of Federal Education and Professional Training in Pakistan launched the National Institute of Excellence in Teacher Education (NIETE) programme, in partnership with Taleemabad, to address a long-standing challenge: how to give thousands of teachers across Islamabad’s public schools access to high-quality, standardised lesson plans without overwhelming them with additional work.
Supported by EdTech Hub’s evaluation, the story of how AI-generated lesson plans were introduced offers critical lessons for education systems navigating the promise and pitfalls of technology.

The NIETE programme rolled out across 341 public schools, reaching 4,044 teachers and nearly 98,000 students. It combined three strands: a digital professional development programme for teachers, AI-generated lesson plans aligned to the national curriculum, and classroom coaching using the TEACH tool. The ambition was clear: to improve instructional quality at scale and reduce the burden of lesson preparation for teachers.

Bridging the Gap

 

The challenge

Before NIETE, lesson planning varied widely. Teachers relied almost exclusively on the textbook, and developing lesson plans of consistent quality was a challenge. This meant students experienced uneven quality in the classroom. NIETE’s introduction of AI-generated plans led to consistent and higher-quality lesson plans, yet some contradictions also emerged. Teachers consistently rated these plans as better than their own, yet platform data revealed low or inconsistent usage. Many teachers felt constrained by the rigid pacing and scripted design, creating a gap between positive perception and actual adoption.

The Intervention

Taleemabad, through NIETE, introduced centrally developed AI-generated lesson plans that followed international pedagogical models such as the Gradual Release of Responsibility and subject-specific approaches like Inquiry-Based Learning for science and the Concrete-Pictorial-Abstract model for mathematics. Teachers accessed these through the NIETE app, while training and mentoring helped them integrate the plans into daily practice. The approach aimed to bring coherence and structure where previously there had been little. EdTech Hub was tasked to evaluate AI-generated lesson plans developed by Taleemabad, an educational technology company in Pakistan, within the framework of the GSMA Innovation Fund for Accelerated Growth and through their implementation in the National Institute of Excellence in Teacher Education (NIETE) programme.

The evaluation aimed to deliver a nuanced understanding of how the centrally developed AI-generated lesson plans can enhance equitable access to high-quality learning content. By triangulating data from diverse methodologies, the findings aim to guide strategic improvements in Taleemabad’s AI-powered lesson plans and NIETE’s deployment of the lesson plans, fostering innovation while addressing systemic educational challenges in LMICs.

The Impact

The evaluation revealed clear benefits. Despite mixed perceptions, the platform data shows that thousands of teachers engaged with the AI-powered lesson plans, with 3,423 teachers accessing at least three. This points to a meaningful shift in adoption, suggesting that the plans were not only accessible but also increasingly integrated into teachers’ practice—helping reduce preparation time and bringing more consistency to classroom instruction.

Yet, the impact was not uniform. Experienced teachers found the scripted design restrictive, and many adapted or abandoned the plans when they clashed with school exam timetables or classroom realities. Issues of differentiation and conceptual depth also limited effectiveness in some contexts.

Further Insights

Lessons on differentiation and conceptual depth

We learnt that: 

  • Teachers saved preparation time and reported that the plans provided consistency and structure.
  • Newer teachers, in particular, valued the scaffolding and step-by-step support.
  • The built-in formative assessments gave educators a stronger sense of student progress.

Core Lessons

Several lessons stand out from our evaluation, including that standardisation brings equity, but flexibility is essential for real adoption. We learn that teachers welcome tools that save time, but they also want the freedom to adapt content to situational needs. Additionally, strong assessment strategies exist, but lesson pacing must align better with local exam calendars. Above all, teacher agency matters. Without space for professional judgment, even high-quality resources risk underuse.

As the use of AI in education becomes more widespread, the findings highlight a clear path forward. Policymakers and EdTech developers must design with teachers, not just for them. That means working collaboratively towards building flexibility into lesson plans, continuing investment in teacher professional development, aligning pacing with assessment schedules and creating feedback loops so teachers shape the evolution of the tools they use.