James Muldoon, Mark Graham & Callum Cant

PACT

* P: Persona

* A: Action (Task)

* C: Context

* T: Target (Format)

PACT is your roadmap to crafting effective prompts.

Clearly define the Persona you want the AI to adopt (e.g., a seasoned scientist, social media manager, a witty comedian).

Then, specify the exact Action you need it to perform (summarize, write, translate).

Next, provide essential Context to guide the AI's understanding (topic, style, audience).

Finally, outline the desired Target format for the output (blog post, script, poem).

By following the PACT prompt framework, you'll significantly improve your chances of getting the AI output you desire.

Here’s an example.

Scenario: A fashion brand wants to launch a new sustainable clothing line and needs engaging social media content.

PACT Prompt:

Persona: You are a passionate environmentalist and fashion influencer.

Action / Task: Create a captivating Instagram caption.

Context: Highlight the eco-friendly materials used in our new clothing line, emphasizing sustainability and style. Target a young, environmentally conscious audience.

Target/ format”: A short, snappy caption with a strong call to action to shop the collection.

As usual finish the prompt with Please ask me any clarification questions?

New in late 2024 this free tool can:

Answer questions and offer key insights on complex documents you upload

Convert complex material into FAQs, briefing documents, audio such as podcasts

Store notebooks of key resources for sharing and as group knowledge bases

• When a user submits a query, first it searches the web for verifiable information, trusted sources

• This search information is combined with the original query to create an enriched prompt

• It uses different AI models including ChatGPT and Clause depending on the task 

• It generates a ‘grounded’ response based on the AI’s knowledge and verified external information

• It cites the sources it used.

• Less prone to hallucination and provides verifiable sources.

By the end of the course all delegates will be able to:

• Understand fundamental AI concepts and terminology.

• Identify practical AI applications for home and business.

• Describe the purpose of a some genAI tools including ChatGPT, Microsoft Co-Pilot, Google Gemini

• Use basic techniques to create a range of documents and communications relevant for own purposes

• Evaluate the potential impact of AI on individuals and society.

• Formulate an action plan for own genAI use and development.

By the end of the course all delegates will be able to:

• Describe the purpose of a range of genAI tools for business including ChatGPT, Microsoft Co-Pilot, Microsoft Translate

• Evaluate the likely benefits and costs of using different versions of genAI and how up-to-date they are

• Use prompt engineering techniques to create a range of business resources relevant for own organisation

• Evaluate some of the disadvantages of genAI use including ethical and data protection issues and the risk of ‘hallucinated’ responses

• Describe likely future developments in genAI

• Formulate an action plan for own genAI use development and organisational development and policy

By the end of the course all delegates will be able to:

• Describe the purpose of a range of genAI tools for business including ChatGPT, Microsoft Co-Pilot, Microsoft Translate

• Evaluate the likely benefits and costs of using different versions of genAI and how up-to-date they are

• Use prompt engineering techniques to create a range of learning and learning support resources relevant for own organisation

• Review examples of AI / digital pedagogy in training and development scenarios

• Evaluate some of the disadvantages of genAI use including ethical and data protection issues and the risk of ‘hallucinated’ responses

• Describe likely future developments in genAI

• Formulate an action plan for own genAI use development and organisational development and policy

Use this guide to assess your current AI governance posture, identify gaps across jurisdictions, and prepare for enforcement-driven scrutiny in 2026.