Human first, AI ready: Why most marketing teams are not
15 Jun 2026
There’s a question unsettling boardrooms and briefing sessions across the UK right now, and it isn’t “which AI tool should we be using?” It’s even harder: “Are we actually in control of what’s happening inside our own marketing organisations?”
As a CMO in 2026, you’re likely somewhere on a spectrum that runs from “kids in the sandpit” – where teams are exploring new tools freely with no strategic framework – to genuinely purposeful, commercially aligned AI deployment. Most organisations, if they’re honest, sit closer to the former than they’d like to admit. And while teams need room to try out new tools – after all, that’s how you learn – “let’s see what happens” can’t be your plan forever.
The token maxing era is over
Until recently, marketing teams have been experimenting with AI consequence-free, testing and iterating with relative carte blanche. But unfettered activity brings all kinds of risk, and we’re already seeing organisations face backlash and reputational fallout over AI-generated content that’s gone out into the world before a human checked whether it should. In fact, Forrester warns that companies will lose over $10 billion globally from ungoverned generative AI use through declining stock prices, legal settlements and fines.
The response from many has been to call for a blanket ban on AI, and while that may feel safe, it just dodges the real leadership problem. The “token maxing era” – that brief, heady window of free experimentation – is ending. What must replace it demands something AI can’t supply on its own: human judgement.
Governance isn’t a brake. It’s a steering wheel.
Marketing people tend to hear “governance” and immediately brace themselves. To some it sounds like slowing down, as if Compliance has joined the chat. But the organisations making the most meaningful progress aren’t rushing to implement more tools, they’re asking the harder questions before the damage happens, like: “Where do the real risks lie?”
In their answers, the same three problem areas keep coming up:
1. Data integrity – what the AI knows (and doesn’t)
So you’ve completely closed off your AI ecosystem, locked it all down to a single tool or private data lake. This may make the AI feel safer – more secure, fewer hallucinations – but it also cuts off its awareness of things happening outside the business, how the market is shifting, how customer language is changing, and what competitors are starting to do differently. When your AI only knows the internal world you’ve fed it, it may give you answers that, although highly confident and technically accurate, are out of date in the real one.
2. Behaviour at the edges – how people use it in practice
Giving people free rein with AI tools without the right guidance gets risky fast, and in marketing it can churn out content that quietly erodes the brand. The AI doesn’t know what it doesn’t know, and crucially, often neither does the person prompting it. Proper governance builds confidence – that’s what turns experimentation into responsible use. But teams must know when an output needs human review before it can leave the building.
3. Cultural disconnect – how are we driving/measuring adoption?
We’re working so fast to avoid being left behind that many organisations have made “AI adoption” an end in itself. “If you’re not using these tools,” they seem to be saying, “you need to explain why not.” Such a fundamental transformation in how an organisation operates must have a clear rationale, both commercially and ethically. If you’re encouraging AI use just to demonstrate uptake and not because it improves the work, you’re measuring activity rather than value.
What AI can do for you – and what it can’t
AI is great for generating options – drafting copy, summarising research, quickly iterating and prototyping ideas – but it should not be allowed to make decisions that define the business, the customer relationship or the brand. Only a human can judge whether something is commercially wise, whether customers will trust it and whether it feels right for the brand.
How to price AI into products and services, for example, is an increasingly common question as work that might have taken weeks in 2022 now takes an afternoon. That’s a business decision – not an AI prompt. When your competitors also use LLMs to generate and iterate content, it’s the human judgement in the room that sets your content apart and identifies it with your brand. And as consumers grow more sceptical about what is real, original or credible, human judgement can steer brands toward winning that trust back.
Buying AI tools is the easy bit, but their value lies in people who know how to use them well, safely and commercially. The Chartered Institute of Marketing’s 2026 European Marketing Agenda found that while 78% of UK CMOs expect budgets to remain stable or grow, money doesn’t automatically translate into using AI well. Even a well-funded CMO can’t make AI work on their own – they need the shared judgement of other teams across technical, legal, commercial and human areas to turn those resources into work that is useful, safe and commercially meaningful.
AI literacy – knowing how to question the tool, guide it and use it responsibly – is what makes those investments pay off. In this context, governance is about building the confidence to move fast without breaking things, not just having control. Marketing leaders must move from asking whether they’re still in control to knowing where it’s human judgement that must lead.
Read more on this topic: The Creative Paradox of AI and the Rise of In-House Agency
