The marketing industry has spent most of this year arguing about which AI platform is best, feverishly trialling new models and salivating at the promise of each new release from the tech giants. It is a brilliant argument. It has rankings, tribes, betrayed loyalties and weekly plot twists. It costs nothing to join, because the moment anyone declares a winner, a new release voids the result.

As sport, it’s fun. As Linkedin bait, chef’s kiss. Comment red herring below for a workflow to solve all your communications challenges. 

As strategy, it’s a decoy.

Because crucially, none of the comparison charts answer the important questions: best at what, for whom, under whose rules, with what consequences? And on those questions, much of the industry has nothing. Not a bad or incomplete answer; no answer at all. 

Crickets. 

The Australian Centre for AI in Marketing quantified this. Of 126 CMOs surveyed, 58% have no documented AI roadmap. Six per cent have a complete one.

The 58% is probably the floor, not a ceiling - leaders rarely under-claim having a plan. More than half of the marketing leaders in Australia considered the question “do you have this written down somewhere?” and said ”yeah, nah.”

That would be alarming even if the only audience asking was the board - but it isn’t.

Australian consumers are already forming their own standards about acceptable AI use. 45% say they would trust a brand less if they discovered its advertising had been created mainly using AI. 86% believe brands should disclose it. And while what people say and what they do are generally two discrete things, there is a rising consumer backlash against the machines. I’d argue AI is more lightning rod than cause, but it has become a proxy for broader fears about deception, dispossession and the erosion of human value.

Contrast those numbers with the roadmap data, and the real problem comes into sharp relief. Consumers increasingly expect brands to explain how AI is being used, while most marketing organisations cannot produce a document explaining where, why or under what conditions they use it.

You cannot disclose what you have not documented.

And you cannot ask consumers to trust a process that nobody inside the business can properly articulate.

The roadmap gap is not merely evidence that marketing is lacking maturity in AI use. It is evidence that brands are making withdrawals from a trust account they are not keeping the books for.

Now the legislation is catching up.

On 15 July, the Prime Minister announced the Office of AI to sit within the Department of the Prime Minister and Cabinet, mandatory Australian AI standards, discussion at National Cabinet next month, and legislation next year. 

Albanese also drew a clear line around Australian creative work, arguing that creators should retain control over how their work is used and valued in AI training. Hard to think of a less ambiguous statement of intent than, “Anything less, is theft”

The details are still to be written, but that is precisely the point.

The political direction is no longer TBC. Creator control, AI training, transparency, and accountability are moving from voluntary guidance into legislation. The government looked at an economy still struggling to define acceptable AI practice, an electorate beginning to resist, and (finally) took a stand. 

The compliance window is the interval before that definition crystallises. Not time left to panic. Time left to exercise leadership.

Writing an AI roadmap is less grand than it sounds. It is not a five-year transformation plan for tools that reshuffle every quarter; not a work of speculative fiction with a 300-page deck and Gantt chart attached.

Start here, start now: start with answers to these questions:

  • What do we use AI for?
  • What will we not use it for?
  • What can we show when somebody asks?

Then build in the hygiene underneath. Which tool contributed to this asset? What data or source material went into it? Was a creator’s work used with permission? What required human review? Who approved the final output?

These are boring questions. They are also the machinery of trust.

A roadmap shouldn’t attempt to predict the future. Maturity ladders call to mind a summit: climb, arrive, stand still, enjoy the view, but in these turbulent times, there is no summit. Less mountain, more ocean. There is this wave, and the next one, the one out back still building, and the quality of your response is the only indicator of outcome.

The roadmap does not predict the waves. It maps the reef.

Consumers are not rejecting AI outright. They are rejecting work that feels deceptive, careless or soulless. They want to know when automation has shaped what they are seeing, and they want - quite reasonably - evidence that somebody remains accountable. They want the human touch not as window dressing, but as proof that judgment still exists somewhere in the system.

For marketers, this should be familiar territory. Trust has always depended on consistency between what a brand says, what it does and what it can substantiate. AI has not changed that rule. It has simply created many more ways to break it at scale.

The legislation and standards will arrive in some form. Consumer expectations already have.

Six per cent of marketing leaders have a complete AI roadmap. Eighty-six per cent of consumers expect disclosure when advertising is mainly AI-generated. The space between those statistics is marketing’s trust crisis in the making: the public is asking questions most organisations are not yet equipped to answer.

The next wave is already building. The question is not which model gets you onto it fastest; it's whether anyone in the organisation has managed to map the reef.