Programme Outline / Workshop Coverage

Module 1: The AI Reality Check

​

Understanding How Generative AI Produces Answers

Participants begin by developing a practical and accessible understanding of how generative AI works.

The focus is not on technical programming knowledge, but on helping employees understand why AI can produce fluent and credible-sounding answers without necessarily understanding whether those answers are accurate.

Participants will explore the gap between how confident an AI response sounds and how reliable it may actually be.

Key Areas Covered

• Introduction to how generative AI works

• How AI generates responses from patterns in data

• Why fluent language can create an impression of authority

• The confidence-versus-accuracy gap

• Strengths of generative AI in workplace applications

• Limitations of AI in professional and leadership contexts

• Situations in which human judgement remains essential

​

Module 2: Ethics and Responsible AI Use

​

Strengthening Responsible Workplace Practices

Employees may use AI tools with good intentions while still exposing the organisation to confidentiality, privacy, security or reputational risks.

This module helps participants understand their responsibilities when using AI in the workplace and the importance of applying appropriate safeguards.

Responsible AI use is positioned not merely as a compliance requirement, but as an essential professional habit.

Key Areas Covered

• Risks associated with workplace use of generative AI

• Confidentiality and sensitive information

• Personal data and privacy considerations

• Intellectual property and ownership concerns

• Hallucinations and inaccurate information

• Bias and potentially discriminatory outputs

• Appropriate human oversight

• Validation and verification processes

• Organisational policies and acceptable-use expectations

• Professional accountability for AI-assisted work

​

Module 3: When AI Gets It Wrong—and How to Catch It

​

Auditing AI-Generated Information

AI errors can be difficult to identify because inaccurate content may be expressed clearly and supported by plausible explanations.

Participants will learn how to break down information before deciding whether it should be accepted, verified or rejected.

They will also practise using structured questions to challenge vague statements, unsupported conclusions and hidden assumptions.

Key Areas Covered

• Distinguishing facts, claims, opinions and assumptions

• Why the distinction matters when evaluating information

• Common types of AI failure

• Hallucinations and fabricated information

• Outdated or incomplete knowledge

• Sycophancy and excessive agreement

• Confirmation bias

• Overgeneralisation

• Unsupported causal relationships

• Missing context and unspoken assumptions

• Precision questioning techniques

• Verifying sources and evidence

​

Module 4: AI as a Thinking Sparring Partner

​

Using AI to Strengthen Rather Than Replace Thinking

AI is often used to provide quick answers. However, it can also be used to challenge assumptions, generate opposing viewpoints and improve the quality of reasoning.

This module introduces critical-thinking disciplines that help participants use AI more deliberately.

Participants will learn how to prompt AI to test their thinking rather than simply reinforce their original position.

Key Areas Covered

• Socratic questioning

• Using questions to uncover reasoning gaps

• Steelmanning an opposing viewpoint

• Differentiating strong counterarguments from superficial disagreement

• Assumption mapping

• Identifying what is being taken for granted

• Prompting AI to challenge a proposal or recommendation

• Asking AI to identify missing stakeholders or risks

• Recognising when AI is merely agreeing with the user

• Testing conclusions against alternative explanations

​​

Module 5: Decision-Making With AI

​

Retaining Human Judgement and Accountability

AI can support analysis and generate recommendations, but it does not remove the need for human ownership of decisions.

Human judgement may be affected by time pressure, overconfidence, familiarity, groupthink and confirmation bias. AI may reinforce these weaknesses if its outputs are not carefully examined.

Participants will explore how to combine AI-generated insights with structured human judgement.

Key Areas Covered

• Common weaknesses in human decision-making

• Cognitive bias and decision shortcuts

• Automation bias and over-reliance on technology

• How AI can amplify existing assumptions

• Pre-mortem thinking

• Imagining failure before implementation

• Identifying risks, consequences and affected stakeholders

• Evaluating the quality and completeness of information

• Applying a practical decision framework

• Determining when escalation or expert review is required

• Maintaining clear human accountability

​

Module 6: The AI Decision Lab

​

Applying Critical Thinking in a Realistic Workplace Challenge

The workshop concludes with an experiential group challenge that brings together the key concepts and frameworks introduced throughout the programme.

Participants work on an AI-assisted scenario involving incomplete information, competing priorities, hidden assumptions and potential risks.

Teams must challenge the AI-generated analysis, develop a defensible recommendation and present their reasoning to their peers.

Capstone Activities

• Reviewing an AI-generated analysis

• Separating facts from assumptions and claims

• Identifying potential hallucinations and bias

• Applying precision questioning

• Mapping assumptions

• Considering credible opposing viewpoints

• Conducting a pre-mortem

• Assessing risks and stakeholder implications

• Developing a recommendation

• Presenting and defending the team’s decision

• Receiving peer and facilitator feedback

​

KEY BENEFITS