AI adoption checklist (Australian Government guidance)

What it does

This skill takes one planned AI tool and checks it against the current Australian Government guidance for safe and responsible AI use. It produces a one-page checklist, an entry for the business's AI register, and a list of open questions for the vendor.

Which guidance is current. The Voluntary AI Safety Standard, with its 10 guardrails, was published on 5 September 2024. On 21 October 2025 the National AI Centre published the Guidance for AI Adoption, which sets out 6 essential practices and, in the department's words, "evolves" the Voluntary AI Safety Standard. The Guidance comes in two versions: "foundations" for businesses starting out or using AI in low-risk ways, and "implementation guidance" for higher-risk or more complex use. The National AI Plan, published 2 December 2025, is a whole-of-government plan and is not a checklist for businesses. This skill uses the foundations version of the Guidance for AI Adoption. Before each use, check the industry.gov.au and ai.gov.au pages below for a newer version.

When to use it

• Before paying for or switching on an AI tool.
• When staff have started using an AI tool on their own and the owner wants to put some rules around it.
• When a customer or partner asks how the business governs its AI use.

What you need

• The name of the tool and what it will be used for, in one sentence.
• Who will use it, and who or what its output affects (staff, customers, job applicants).
• What information will go into it, especially any personal or customer information.
• The vendor's terms, privacy policy or data handling page, if available.
• Optional: Google Workspace or Microsoft 365 connector, to save the checklist and register entry as a document. Not required.

Steps for Claude

1. Restate the use in one sentence and ask any missing questions from the list above.
2. Screen the risk. Ask: does the tool make or influence decisions about people, talk to customers directly, handle personal or sensitive information, or run without a person checking its output? If the answer to any of these is yes, say so and recommend reading the implementation guidance as well as the foundations guidance.
3. Walk through the six essential practices and turn each into two or three yes or no checks for this tool:
   1. Decide who is accountable. Is one named person responsible for this tool? Does the business have an AI policy, even a short one?
   2. Understand impacts and plan accordingly. Who could be affected, including vulnerable people? How can someone raise a problem or challenge an outcome?
   3. Measure and manage risks. Has the use been screened for risks using the government's screening questions? What would make the business stop using it?
   4. Share essential information. Is the tool recorded in an AI register? Will customers know when they are dealing with AI, for example a chatbot?
   5. Test and monitor. Has the vendor shown evidence of testing? Has the business tried it on real examples before relying on it? Who checks it over time? Do the business's data and cyber security practices cover it?
   6. Maintain human control. Where does a person review output before it goes out? How can the tool be paused or switched off, and what is the fallback?
4. Mark each check Done, To do or Not applicable, with a one-line reason.
5. Draft an AI register entry and a list of questions for the vendor.
6. If personal information is involved, recommend running the au-privacy-ai-check skill as well.

Output format

1. Risk screen result: low, needs attention, or high, with reasons.
2. The checklist, grouped under the six practices, with status and owner.
3. AI register entry: tool, purpose, owner, data used, who is affected, review date.
4. Vendor questions.
5. The three most important actions before go-live.

Australian rules and sources

• Voluntary AI Safety Standard (published 5 September 2024, page updated 2 December 2025, with a notice pointing to the newer guidance): https://www.industry.gov.au/publications/voluntary-ai-safety-standard
• Guidance for AI Adoption (published 21 October 2025; opens on ai.gov.au): https://www.industry.gov.au/publications/guidance-for-ai-adoption
• Guidance for AI adoption: foundations: https://www.ai.gov.au/staying-safe-and-responsible/essential-ai-practices/guidance-ai-adoption-foundations
• AI screening questions: https://www.ai.gov.au/practical-guides-and-learning/planning-tools-and-templates/ai-screening-questions
• AI policy guide and template: https://www.ai.gov.au/staying-safe-and-responsible/essential-ai-practices/create-ai-policy
• AI register template: https://www.ai.gov.au/staying-safe-and-responsible/essential-ai-practices/ai-systems-register
• National AI Plan (2 December 2025): https://www.industry.gov.au/publications/national-ai-plan

Limits

The Guidance for AI Adoption is voluntary. Existing laws still apply to AI use, including privacy law, the Australian Consumer Law, anti-discrimination law and negligence, as the department sets out in "The legal landscape for AI in Australia": https://www.industry.gov.au/publications/voluntary-ai-safety-standard/legal-landscape-ai-australia. This checklist is general information, not legal advice. For uses that affect people's jobs, money, health or access to services, get professional advice.

Example requests

• "We want to add an AI chatbot to our website to answer booking questions. Run the government checklist on it."
• "Half my team is already using ChatGPT for quotes. Give me a checklist and a simple register entry."
• "Is the Voluntary AI Safety Standard still the thing to follow, and what do I need to do for our new AI note-taker?"