01
Overview
A practical, critical grounding in AI: how models work, where they fail, and how to deploy them accountably. Includes hands-on building with hosted models and a governance framework for real use.
02
Aims of the Course
- Explain how modern AI systems are trained and used
- Identify bias, hallucination and reliability limits
- Build a working AI-assisted workflow or tool
- Write an accountable AI use policy
03
Course Structure
01
How AI Works
- Machine learning and neural networks
- Large language and multimodal models
- Training data and its consequences
- What models cannot do
02
Prompting and Evaluation
- Structured prompting
- System instructions and constraints
- Testing outputs against criteria
- Hallucination detection
03
Building With AI
- Calling model APIs
- Retrieval over your own documents
- Automating a repetitive workflow
- Cost and latency trade-offs
04
Bias, Safety and Harm
- Sources of bias
- Privacy and data minimisation
- Deepfakes and synthetic media
- Accessibility and exclusion
05
Governance and Accountability
- Human oversight and escalation
- Transparency and disclosure
- Regulatory landscape
- Writing an internal AI policy
06
Applied Project
- Scoping a responsible AI use case
- Prototype development
- Evaluation and risk log
- Presentation and deployment plan
04
Learning Outcomes
- 01Explain model capabilities and limits accurately
- 02Design prompts that produce reliable output
- 03Build a small AI-assisted workflow
- 04Assess an AI use case for bias and privacy risk
- 05Write an organisation-ready AI use policy
05
Assessment
- Working AI prototype
- Risk and evaluation report
- AI policy document
06
Entry Requirements
- Digital literacy; basic spreadsheet or coding comfort helpful
07
Progression and Careers
- AI operations specialist
- Innovation and automation lead
- Responsible AI advisor