Programme overview
Technical depth: Practitioner · Practical mode: Case study
Introduction
Government entities are under pressure to use generative AI to answer citizens faster, draft policy material and reduce administrative workload, yet public bodies carry duties of fairness, transparency and accountability that private firms do not. Poorly governed deployments risk wrong answers to the public, leaked information and decisions that cannot be explained or contested. This Core Concept course equips public sector managers to select, design and oversee generative AI use cases against international principles and standards. Participants work on cases from several government functions and leave with a Responsible Generative AI Deployment Plan for one of their own services.
Course Objectives
- Assess where generative AI can add public value across service delivery, policy work and internal operations
- Apply international trustworthy AI principles and standards to classify and screen government AI use cases
- Design generative AI services, including knowledge assistants and citizen-facing chatbots, with human oversight and accessibility built in
- Conduct AI system impact assessments and prepare transparency records for public disclosure
- Specify security, privacy and procurement controls for generative AI solutions and model providers
- Produce a Responsible Generative AI Deployment Plan ready for review by the entity's leadership
Target Audience
- Service delivery and customer experience managers in ministries, agencies and municipalities
- Digital transformation and innovation managers leading AI initiatives in public bodies
- Policy, legal and regulatory affairs managers who draft and review official material
- Data, information management and cybersecurity managers responsible for government information assets
- Procurement and contract managers who buy AI solutions and cloud services for public entities
Course Outline
Day 1: Generative AI and Public Value
- Public Sector AI Use Cases Across Service Delivery, Policy and Internal Operations
- Generative AI Capabilities and Limits: Large Language Models, Retrieval-Augmented Generation and Agents
- Public Value Framework for Assessing AI in Government Services
- OECD Framework for Trustworthy AI in Government: Enablers, Guardrails and Engagement
- Entity AI Readiness Assessment Across Data, Skills, Infrastructure and Policy
Day 2: Principles, Standards and Classification
- OECD AI Principles and the UNESCO Recommendation on the Ethics of AI
- OECD Framework for the Classification of AI Systems
- ISO/IEC 42001 AI Management System Controls for Public Entities
- NIST AI 600-1 Generative AI Profile Risk Categories
- Government Data Classification and Hosting Options for AI Workloads
Day 3: Designing Responsible Generative AI Services
- Generative AI Use Case Screening Matrix for Government Services
- Retrieval-Augmented Generation Design for Policy and Regulation Knowledge Assistants
- Citizen-Facing Chatbot Design with Human Handover and WCAG 2.2 Accessibility
- Prompt and Output Guidelines for Drafting Official Correspondence
- Procurement Requirements for Generative AI Solutions and Model Providers
Day 4: Impact, Transparency and Security Risks
- ISO/IEC 42005 AI System Impact Assessment for Public Services
- Algorithmic Transparency Record and Public Disclosure Template
- Multilingual Accuracy and Bias Testing of Generative Outputs
- Prompt Injection and Data Leakage Controls from the OWASP Top 10 for LLM Applications
- Human Oversight and Contestability Mechanisms for AI-Assisted Decisions
Day 5: Case Work and the Deployment Plan
- Social Benefits Case Study: Reviewing an AI-Assisted Eligibility Process
- Government Contact Centre Case Study: Rolling Out a Generative AI Assistant
- Responsible Generative AI Deployment Plan Drafting for an Own Service
- Red-Team Review of the Plan Using Misuse Scenarios
- Leadership Panel Presentation and Plan Defence
Skills You Will Gain
- Public Value Assessment
- Trustworthy AI Principles Application
- AI Use Case Screening
- Generative AI Service Design
- AI Impact Assessment
- Algorithmic Transparency Reporting
- Generative AI Security Controls
- AI Procurement Specification
Why Attend This Course
- Return to work with a Responsible Generative AI Deployment Plan for a real service, already tested against misuse scenarios
- Explain to leadership and the public how an AI-assisted service works, where humans stay in control and how decisions can be challenged
- Spot the design and procurement choices that expose an entity to inaccurate answers or leaked information
- Compare approaches with public sector managers from other countries and levels of government
Conclusion
Generative AI can make public services faster and more accessible, but only when public bodies keep fairness, transparency and accountability at the centre of every deployment. This course moves from public value and readiness, through the international principles and standards for trustworthy AI, to service design, impact assessment, transparency and security controls. The final day turns that material into a Responsible Generative AI Deployment Plan that participants take back to their entity for leadership review.