Let's address the elephant in the room. Yes, we use AI at Cohen Media House. We use it extensively. We use it for ideation, for production workflows, for content iteration, for research, and for a dozen other tasks that used to eat hours of our week. We are not apologetic about this. But we are very deliberate about how we use it — and more importantly, where we don't.
Because here's what most agencies won't tell you: the difference between using AI well and using AI badly isn't a question of technology. It's a question of creative leadership. AI is a power tool. In the hands of someone with vision, it's extraordinary. In the hands of someone without it, it produces the most sophisticated mediocrity you've ever seen.
The AI Landscape in Creative Production
As of early 2026, the creative industry is roughly split into three camps. The first camp refuses to use AI at all, either on principle or out of fear, and is watching their competitors produce three times the output at half the cost. The second camp has gone all-in on AI, automating everything from copywriting to image generation, and is producing a flood of content that looks technically proficient and feels entirely soulless. The third camp — the one we're in — is using AI as an accelerant for human creativity, not a replacement for it.
The third camp is the smallest. But it's producing the best work.
Where AI Excels (And Where We Let It)
We're transparent with our clients about exactly where AI fits into our process. Here's where it works brilliantly:
Research and ideation. Before a brand strategy session, AI can synthesise competitor analysis, audience research, and trend data in minutes instead of days. This doesn't replace strategic thinking — it accelerates the information-gathering that precedes it. We walk into every strategy session better informed than we would have been otherwise.
First-draft copy iteration. When we're developing messaging for a campaign, AI can produce fifty variations of a headline in seconds. Ninety percent of them will be forgettable. But the exercise of seeing all those variations — and understanding why they don't work — sharpens our thinking about what will. It's like a sparring partner. It doesn't fight the bout for you, but it makes you sharper for the real thing.
Production workflows. Background removal, image resizing, format adaptation, basic colour correction, transcription, subtitle generation — the mechanical tasks of content production that used to consume hours of a skilled editor's time. AI handles these with near-perfect accuracy, freeing our team to spend their time on creative decisions instead of technical grunt work.
Content repurposing. Taking a long-form video and identifying the best clips for social, adapting a blog post into social captions, reformatting assets across platforms — AI is exceptional at this kind of systematic adaptation. What used to take a day now takes an hour.
Where AI Fails (And Where We Refuse to Use It)
Here's the part most agencies skip. The limitations are just as important as the capabilities, and ignoring them is how you end up with content that's technically impressive and emotionally empty.
Creative direction. The most important decision in any creative project is not what to make, but what not to make. It's the ability to look at ten options and identify the one that's right — not just good, but right for this brand, this audience, this moment. AI cannot do this. It can generate options. It cannot exercise taste. Taste is the product of lived experience, cultural understanding, and aesthetic conviction. It's the reason you hire a creative director, and no language model will replace it in our lifetime.
Brand voice. AI can mimic a tone of voice. It cannot create one. The difference is subtle but critical. A brand voice that was developed by AI will always feel like a composite — a statistical average of everything it's been trained on. A brand voice developed by a human who understands the brand's personality, audience, and aspirations will feel specific, distinctive, and true. We develop every brand voice by hand. AI doesn't touch it.
Photography and film direction. There are agencies generating campaign imagery entirely with AI. Some of it looks stunning. None of it looks real. And for the industries we serve — hospitality, tourism, fashion — authenticity isn't a nice-to-have. It's the entire point. When a guest books a hotel, they need to trust that what they're seeing represents reality. AI-generated imagery, no matter how beautiful, undermines that trust at a fundamental level.
AI is the best assistant in the world. But it's a terrible boss. The moment you let it make creative decisions instead of inform them, the work suffers.
Our Framework: The 80/20 Principle
Here's the simple model we use to decide where AI belongs in any project:
If a task is 80 percent mechanical and 20 percent creative — background removal, format adaptation, data synthesis — AI leads and a human reviews.
If a task is 80 percent creative and 20 percent mechanical — brand strategy, creative direction, photography, campaign concepts — a human leads and AI assists where useful.
No task is ever 100 percent AI. There is always a human checkpoint. Always. Because the thing that separates good work from great work is judgment, and judgment is the one thing AI fundamentally cannot replicate.
The Client Conversation
We've found that the best approach with clients is total transparency. We tell them exactly where AI is in our workflow, exactly what it's doing, and exactly where the human decisions are. Most clients are relieved by this conversation, because they've been hearing conflicting messages from the industry — some agencies claiming AI is evil, others claiming it does everything, and very few being honest about the nuance.
The question isn't whether to use AI. That ship has sailed. The question is whether the humans guiding the AI have the creative judgment to use it well. That's the value proposition of working with a team that has twelve years of real-world experience in the industries they serve. The AI is new. The judgment isn't.
What Comes Next
AI tools are improving at a staggering rate. Capabilities that seemed impossible twelve months ago are now routine. This trend will continue. But here's what won't change: audiences can tell the difference between content that was made with care and content that was assembled by an algorithm. That instinct is hardwired.
The agencies that thrive in this era won't be the ones that use the most AI or the least. They'll be the ones that use it with the most intention — that understand where the technology ends and the craft begins, and that refuse to blur that line for the sake of speed or margin.
That's our commitment. The tools evolve. The standard doesn't.