Communication and Public Relations Australia has launched a 12-month AI Working Group to develop ethical principles and practical frameworks for the profession, as the federal government stands up an Office of AI. The signal for marketers: set your own rules now.
The rules are coming either way. You either write your own version first or you inherit someone else's under pressure.
Communication and Public Relations Australia has launched an AI Working Group to help the comms and PR profession get ahead of how fast the technology is being adopted. Over the next 12 months the group will look at AI across the whole career lifecycle, from students entering the field to executives making decisions, and produce ethical principles, practical frameworks and policy positions for the CPRA board.
The timing is not an accident. The federal government has stood up an Office of AI inside the Department of the Prime Minister and Cabinet and is developing Australian Standards for AI. The industry is moving to shape its own rules before the national ones land.
That instinct is the right one. The organisations writing their own principles now are the ones who will not be scrambling to comply later.
Why it matters
Most marketing and comms teams are already using AI daily, for drafting, research, images and analysis, and most are doing it with no agreed internal rules at all. That gap is where the risk sits. Who checks the output, what data goes into the tools, what gets disclosed, who is accountable when it goes wrong.
A national framework will eventually set a floor. Waiting for it is the flying-blind option. The businesses that decide how they will use AI now, on their own terms, get to build good habits before compliance forces bad ones on them.
The window CPRA has given its AI Working Group to develop principles and frameworks, as Australia builds national AI standards in parallel.
What to do about it
Write down how your team uses AI today. You cannot set rules for something you have not mapped. Start with what is actually happening.
Agree the basics in one page. What tools are approved, what data never goes into them, what gets a human check, what gets disclosed. Keep it short enough that people read it.
Name who is accountable. AI does not remove responsibility for the output. Someone owns what ships, the same as always.
Watch the national standards taking shape. The Office of AI and Australian Standards will set expectations. Track them so your rules line up rather than clash.
Revisit it as the tools change. This is not a policy you write once and file. The capability moves, so the rules should be reviewed, not set in stone.
AI governance sounds like a big-company concern. It is not. A one-page set of principles you actually follow beats a thick policy nobody reads, and it beats having no answer when a client or a regulator asks how you use these tools.