Using AI to strengthen extension outcomes
December 11, 2025 | Demi Taylor
Using AI to strengthen extension outcomes
As the Ag Ex and SA Drought Hub MEL Officer, most of my work centres around gathering data and stories from Hub partners across the state, so we can better understand practice change, adoption, and what projects are making a real difference. And like many, I’m always looking for ways to do things more efficiently, particularly during peak seasons.
That’s why I signed up for the Using AI in Extension workshop delivered by Denise Bewsell and John James from Enablers of Change. I went in curious and left genuinely excited about how much time and clarity AI can offer farming systems groups and other agricultural organisations.
The good news is, AI isn’t here to replace us
The presenters began the workshop with a reminder that AI is not a replacement for people; it’s a tool to save on time, enhance communication, support decision-making, and spark new ideas.
That message really stuck with me. Relationships, trust, and local context matter in extension, and AI proves valuable is freeing up time so you can focus on people, not paperwork.
For drought resilience work in particular, any tool that can help interpret data quicker, communicate information more effectively, and share insights in ways that support greater adoption can make a real difference on the ground.
When AI is useful, and when it’s not
Artificial intelligence shows up in different ways in everyday life, and the workshop helped clarify how the different terms fit together. AI is the broad umbrella term, covering familiar tools like voice assistance, predictive text, and recommendation systems. Within that sits machine learning, which helps systems learn from data, and powers things like spam filters and image classification. A more recent development is generative AI, which can create new content such as artwork, videos, or music. And finally, large language models (LLMs) (like ChatGPT) are a type of machine-learning model that are able to answer questions, summarise, information, and support interactive chatbots.
One of my favourite visuals from the workshop was a ‘Using Generative AI’ risk grid, to better support decision-making of when to utilise these systems/tools.
The grid makes it easy to determine when AI is appropriate, and I now think of it like this:
- If the task is low risk and time-consuming= consider using AI
- If the task is high risk or accuracy is critical= stick to usual methods
It’s a great way to help extension professionals and project teams feel confident about giving AI a go, without worrying about misusing it or becoming reliant on it.
How AI can help farming groups
The workshop highlighted how AI can assist with extension outcomes. Farming systems groups can use AI to assist with:
1. Reporting, monitoring, and evaluation
- Summarise qualitative survey responses
- Process quantitative data efficiently
- Identify common themes
- Draft structures or sections of project reports
- Turn complex insights into simple key messages
- Determine consensus on an issue- using tools like Consensus AI to scan research databases and identify what the evidence actually says
2. Designing extension events
- Writing run sheets
- Mapping out agendas
- Creating activity ideas
- Drafting facilitator notes
3. Communications and extension materials
- Social media posts
- Flyers or event advertisements
- Summaries of technical information
- Tailored messages for different audiences
One of the examples shared was using AI to turn lengthier technical reports into short and actionable summaries- a simple solution for getting actionable key messages to busy growers who don’t have the time to be reading full reports!
4. Generating new ideas
When you don’t know where to start with a task or project, AI can help with brainstorming, draft structures or frameworks, reframing ideas, or offering alternative approaches. As John James described it; “it’s like having an overly enthusiastic intern”.
Writing better prompts
The ‘prompt architecture’ shared by Denise and John was a game-changer for me:
- Task– what needs to be created from the prompt (write, summarise, analyse, brainstorm, word count etc.)
- Context– background to the situation/task (industry, region, target audience)
- Examples– what does ‘good’ look like? (upload good examples as a basis of what needs to be created)
- Persona– what role should it play? (e.g., podcast writer, analyst, consultant speaking to a farmer, devil’s advocate)
- Format– language and structure (Australian language, dot points vs narrative)
- Tone– how should it sound? (formal, professional, chatty, humorous)
- Check (prompting questions, and check for hallucinations)
Since the workshop, I’ve tried incorporating the ‘check’ component, by using the prompt both presenters kindly shared with us; “before you begin, ask me questions to be 95% confident of doing an excellent job”. It prompts AI to ask for more context and genuinely improves the output!
Why this matters for drought resilience
For me, the biggest takeaway was that small efficiencies add up.
If AI helps your organisation to:
- Communicate more clearly
- Analyse information faster
- Plan events more efficiently and,
- Report on impact in record time
…then you’re freeing up time to focus on the real work of supporting farmers to make informed, confident, and resilient decisions.
Ethical and responsible use of AI
It’s important to note that the workshop also covered the ethical and responsible use of AI. This included strong reminders about protecting confidential information by avoiding the use of public tools for sensitive data and ensuring data training is turned off where required. We were also encouraged to double-check all AI-generated content, as AI can sometimes produce inaccurate information, and to view AI as a tool for drafting, structuring, and idea generation rather than a source of authoritative facts. Above all, AI should support (not replace) the human warmth and authenticity central to agricultural extension.
Final reflections and acknowledgement of content
This was a brilliant workshop, and I would encourage other extension professionals to get involved in similar sessions or explore additional AI learning opportunities as this field continues to evolve.
My key takeaway from both this workshop and David Warren’s keynote presentation at the 2025 APEN Conference is that organisations not engaging with or considering AI in some capacity risk being left behind. Even small, careful uses of AI can help organisations strengthen their visibility, improve search engine optimisation, better understand their digital audience and relationships, and ensure extension content actually reaches the people who need it most (our growers).
The insights shared in this article are drawn from content presented throughout the workshop. You can learn more about Denise and John via Enablers of Change.
You can also subscribe to their free weekly newsletter here or read or listen to ‘Is AI going to steal our jobs?’.
Disclaimer:
This article was written from the individual perspective of the Ag Ex and SA Drought Hub Monitoring, Evaluation, Reporting, Learning and Adoption Officer, Demi Taylor.
Any recommendations, suggestions or opinions contained in this publication do not necessarily represent the policy or views of Ag Excellence Alliance or the South Australian Drought Resilience Adoption and Innovation Hub (SA Drought Hub). No person should act on the basis of the contents of this publication without first obtaining specific, independent, professional advice. Ag Excellence Alliance and the SA Drought Hub and contributors to this publication may identify products by proprietary or trade names to help readers identify particular types of products. Ag Ex, the Hub, and contributors do not endorse or recommend the products of any manufacturer referred to. Other products may perform as well as or better than those specifically referred to. Ag Excellence Alliance and the SA Drought Hub will not be liable for any loss, damage, cost or expense incurred or arising by reason of any person using or relying on the information in this publication.