IAB Australia has released a guide titled "Prompting with LLMs" aimed at marketing professionals navigating AI tools in their daily workflows. The guide covers prompting fundamentals, structured techniques for different marketing use cases, and risk considerations that most vendor-produced guides conveniently skip.
The document is notable for what it is not. It is not a breathless endorsement of AI replacing marketers. It is not a vendor pitch dressed up as education. It is a practical framework that acknowledges both the utility and the limitations of current LLM technology in a marketing context.
The guide covers four core prompting techniques: role-based prompting (assigning the LLM a specific persona or expertise), chain-of-thought (breaking complex tasks into sequential steps), few-shot examples (providing sample outputs to calibrate quality), and constraint-based prompting (setting explicit boundaries on tone, length and format). These are not new concepts in the AI space, but packaging them for a marketing audience with relevant examples is genuinely useful.
What sets it apart is the risk section. The guide explicitly addresses hallucination risks, copyright considerations for AI-generated content, data privacy obligations under the Australian Privacy Act, and the need for human review before any AI output reaches a customer. These are the conversations that matter and that most "how to use AI" content skips entirely.
Of Australian marketers report using AI tools at least weekly, but only 23% have formal usage guidelines (IAB Australia, 2026)
The gap between adoption and governance is the real story. Two-thirds of marketers are using AI tools regularly. Fewer than a quarter have any formal policy around how those tools should be used, what data can be input, or who reviews the output. That is a compliance problem waiting to surface.
IAB Australia has positioned itself as the industry body willing to do the unglamorous work of building frameworks. Their previous guides on programmatic supply chain transparency and retail media measurement followed a similar pattern: practical, balanced, occasionally boring in exactly the right way.
Why it matters
Marketing teams are adopting AI tools faster than their organisations are building guardrails. The IAB guide gives teams a shared vocabulary and a baseline framework that can be adapted to their specific context. It is particularly useful for mid-market businesses that do not have dedicated AI or legal teams to develop internal policies from scratch.
What to do about it
Download the guide from IAB Australia's website. Use it as a starting point for an internal AI usage policy. Focus on three things: what data can and cannot be entered into LLM tools, who reviews AI-generated content before it is published, and how you are documenting AI use for compliance purposes. The prompting techniques are helpful. The governance framework is essential.
