Thornhill College in Derry, Northern Ireland, has won the post-primary category at the 2025–26 Northern Ireland Schools ICT Excellence Awards for its innovative use of both Microsoft Copilot 365 and Google Gemini Pro in teaching and learning. The award, which celebrates outstanding integration of technology in the classroom, highlights a growing trend—schools moving beyond single-platform AI deployments toward a best-of-both-worlds approach.
The recognition comes as schools worldwide grapple with how to safely and effectively bring generative AI into education. Thornhill’s model, praised by judges for its “sustained, whole-school commitment” to AI, offers a concrete blueprint for educators and IT leaders navigating this new terrain.
The Winning Formula: Inside Thornhill’s Dual-AI Approach
Thornhill didn’t simply sprinkle AI into a few lessons. According to the award announcement, the school embedded both tools deeply into daily workflows. Microsoft Copilot 365—the enterprise-grade assistant integrated into Word, Excel, PowerPoint, and Teams—was deployed for staff productivity. Teachers used it to generate differentiated lesson plans, craft feedback rubrics, and summarize long email threads, reclaiming hours of administrative time each week.
Meanwhile, Google Gemini Pro took center stage in student-facing activities. In English classes, learners collaborated with Gemini to brainstorm essay structures and receive instant tips on grammar and style. In science, they used it to simulate research scenarios, querying the model for explanations of complex concepts and then fact-checking outputs—a critical skill in an age of AI hallucinations. The dual approach, the judges noted, allowed the school to leverage Copilot’s deep Office integration for institutional efficiency while tapping Gemini’s broad creative potential for active learning.
Crucially, the rollout was underpinned by a rigorous ethical framework. Thornhill established clear policies on age-appropriate use, data privacy, and academic integrity. Teachers received extensive professional development not just on how to use the tools, but on how to redesign assessments so that students learn with AI rather than being replaced by it.
What the Award Means for Different Stakeholders
For Educators and School Leaders
This award signals that AI in education has moved from pilot experiment to practical reality. Thornhill’s success demonstrates that a thoughtful, multi-vendor AI strategy can work in a real secondary school setting. The key takeaway: start by solving real pain points—like teacher workload—before expanding into student-facing applications. Don’t lock yourself into one ecosystem prematurely. The school’s decision to use Copilot for staff and Gemini for students shows that interoperability is not only possible but may yield better outcomes.
For IT Administrators
Managing two cloud-based AI platforms raises non-trivial challenges, from identity management to data governance. Copilot for Microsoft 365 typically requires education customers to have A3 or A5 licenses, with the AI add-on available at a per-user cost. Gemini Pro, on the other hand, might be accessed through Google Workspace for Education Plus or the Teaching and Learning Upgrade, or via Google AI Studio for more advanced use. Administrators must carefully configure both environments to prevent accidental data leakage—for example, by ensuring that student records stored in OneDrive aren’t inadvertently exposed to Gemini queries. Thornhill’s IT team reportedly set up dedicated security groups, turned off history retention where appropriate, and trained staff on data classification. Such steps are now rapidly becoming baseline requirements for any school introducing generative AI.
For Parents and Students
There’s natural anxiety about AI in classrooms, but this award provides reassurance. Thornhill’s clear policies mean students are explicitly taught how to use AI ethically—as a tutor, not a shortcut. Parents can see a model where technology augments teachers rather than replacing human connection. The school’s communication emphasized that AI outputs are always reviewed, never blindly trusted, reinforcing critical thinking. That transparency is essential for building community trust.
From Experiment to Excellence: The Short History of AI in Education
The journey to this point was rapid. Before late 2022, AI in schools mostly meant rule-based tutoring systems or basic automated essay scoring. The debut of ChatGPT changed everything, sparking both excitement and panic. Microsoft moved quickly to integrate OpenAI’s models into its products, launching Copilot for Microsoft 365 in early 2023. Google responded with Bard, later evolved into Gemini, and began weaving it into Workspace and classroom tools.
Northern Ireland’s education system, guided by its EdTech strategy, encouraged controlled experimentation. The annual ICT Excellence Awards, run by the Education Authority, became a spotlight for schools pushing boundaries. Previous winners had showcased coding clubs and tablet rollouts, but this cycle saw a sharp turn toward generative AI. Thornhill was among a handful of institutions that boldly adopted not one but two major AI platforms, a move that required navigating different privacy regimes and licensing models.
The school’s own timeline likely began with a teacher-led pilot in early 2024, followed by cross-departmental buy-in, formal professional learning communities, and iterative policy development. By the 2025–26 school year, the dual-AI model was woven into the fabric of the school, attracting the attention of inspectors and award committees.
Your Action Plan for AI Adoption in Schools
Drawing on Thornhill’s playbook and broader best practices, here’s a practical roadmap for school leaders and IT departments ready to embark on their own AI journey.
-
Assess Readiness and Set Objectives
Survey staff digital confidence, audit existing infrastructure, and define what you want AI to achieve—whether it’s reducing teacher out-of-hours work, personalizing learning, or boosting creative thinking. Don’t adopt AI for its own sake. -
Choose Your Stack Wisely
- If your school is a Microsoft 365 house, Copilot’s integration is unbeatable for teacher productivity. Licenses require an A3/A5 base.
- If you’re on Google Workspace, Gemini offers natural language processing for student collaboration. The Teaching and Learning Upgrade adds AI features to Classroom and Docs.
- Consider Thornhill’s hybrid model: many schools already use both ecosystems; AI can coexist if managed correctly. -
Start with a Staff Pilot
Equip a small cohort of tech-savvy teachers with AI tools, provide training, and collect feedback. Use their experiences to craft use-case guides. Only after staff are confident should you introduce student-facing AI, and then only with careful scaffolding. -
Build an Ethical and Privacy Framework
Address data protection (GDPR in NI), bias, misinformation, and academic honesty head-on. Publish a simple, accessible acceptable-use policy. Configure AI settings to minimize data retention—Copilot and Gemini both offer admin controls. -
Invest in Professional Development
This is the single most important factor. Training must go beyond button-clicking to cover lesson redesign, prompt engineering, and ethical discussions. Thornhill’s success was built on sustained, collegial professional learning, not a one-off workshop. -
Evaluate and Iterate
Set measurable goals: reduced marking time, improved student engagement metrics, higher-quality project work. Regularly review the initiative’s impact and adjust. Engage parents and governors through update sessions.
The Road Ahead
Thornhill’s award is not an endpoint. Microsoft is set to roll out Copilot “agents”—autonomous assistants that can perform multistep tasks—into education tenants later this year. Google’s Gemini is gaining screen-reading capabilities through Project Mariner, potentially offering real-time coaching during student work. Meanwhile, the EU’s AI Act and UK-specific regulations will add new compliance layers that schools must anticipate.
The school’s model will likely evolve, perhaps toward personalized AI tutors that adapt to each student’s pace, or cross-curricular projects co-designed with large language models. But the foundational principles—transparency, teacher empowerment, and ethical rigor—will remain. For any school watching, the message is clear: the AI era in education is no longer a distant speculation. It’s here, it’s award-winning, and it’s replicable.