For as long as agencies have existed, clients have been told the same thing: you can have good, fast or cheap, but you can only pick two. It was a fair summary of the economics. Skilled human time was scarce, so quality, speed and cost pulled against each other.
AI has changed that. James Poulter, in his book AI at Work, argues that focused, task-specific AI tools now challenge the old triangle, and we think he’s right. When research, drafting and production can happen in minutes, speed and cost stop being things you trade quality against. They become the baseline. Every agency, in-house team and freelancer now has access to the same tools, so being fast and cheap is no longer a reason to hire anyone.
That sounds like good news for clients, and in many ways it is. But there’s a catch.
AI makes competent work cheap. It doesn’t make good work cheap.
AI is very good at producing content that looks right: fluent, on-brand-ish and grammatically clean. What it can’t do by itself is decide what a brand should say, what it should leave out, or what it shouldn’t risk. Left unsupervised, it drifts towards the average of everything it has learned. The result is marketing that is technically polished and strategically empty, produced faster than ever.
Speed multiplies whatever is put in. Good thinking gets better and faster. Weak thinking just gets published sooner.
Why this matters most in regulated sectors
In financial services, legal and professional services, healthcare, veterinary and animal healthcare and other regulated fields, the stakes are higher than a forgettable campaign. Marketing has to meet strict standards. Financial promotions must be fair, clear and not misleading. Advertising must comply with the ASA and CAP codes. Professional regulators set their own expectations on what can be claimed and how.
AI has no understanding of any of that. It can produce a confident claim that isn’t substantiated, or a persuasive line that crosses a regulatory boundary, and it will do so in exactly the same tone as everything else. It cannot be held accountable to a regulator. Your organisation can, and so can the people it works with.
That’s why we don’t see AI as a replacement for expertise, but as something that raises the value of it. When output is abundant, the scarce and valuable thing is the judgment to know what’s right, what’s compliant and what’s true.
What good looks like now
We think good marketing in the AI era passes three tests. They’re the ones we apply to our own work.
Is it ownable? Could a competitor run the same campaign just as well? If so, it isn’t yours, however polished it is. Ownable work comes from a real understanding of what makes a client different, not from a generic prompt.
Does it show judgment? Good work carries the fingerprints of real decisions: something cut, a risk considered, an option deliberately rejected. In regulated sectors, that includes knowing where the lines are and staying well inside them.
Does it have provenance? Can you trace why a piece of work exists and what it says back to something true and specific about your organisation? Every claim should have a source, every message a reason, and every piece of content a person who checked it and stands behind it. That’s what protects you when a regulator, a customer or a journalist asks how it came to be.
Where an agency like Mobas fits
If an agency’s value was mostly production capacity, AI is a serious problem for it. That was never really the point of a good agency. The point is strategic judgment, deep knowledge of a client’s sector and market, and the accountability that comes from a long-term working relationship.
We use AI throughout our work: for research, for analysis, for testing ideas and for taking on the heavy lifting that used to slow everything down. It means we can spend more of our time on the parts that matter, like strategy, creative direction and making sure everything is accurate and compliant. But every output is checked, shaped and owned by people who understand your business and the rules you operate under.
Used well, AI means clients don’t have to choose between speed, cost and quality. Used badly, it means all three look fine until something goes wrong. The difference is the people behind it.
If you’d like to talk about how AI could work for your organisation, without compromising on quality or compliance, we’d be glad to hear from you.
