AI hasn't replaced the creative director. It hasn't eliminated post-production. And it definitely hasn't resolved copyright. Here's what's actually true...
If you've sat in a pitch recently, you've heard some version of the same claim: AI has changed everything. Work that used to take weeks now takes hours. Visuals that required a full production team can now be conjured from a text box. The creative industry has been democratised, disrupted, and fundamentally transformed. Some of this is true. Most of it is oversimplified to the point of being misleading.
We've been using andtesting AI tools at Pennybridge to find a clear point of view about where they genuinely add value and where they don't. What we've found over sustained real-world use is that the gap between what AI is advertised to do and what it actually takes to produce professional creative output is significant. And until the industry has an honest conversation about that gap, clients will keep arriving with unrealistic expectations, budgets calibrated to a fantasy, and frustration when reality doesn't match the demo reel.
So here's what's actually true.
Myth 01
"One prompt and it's done — AI makes great creative instantly."
The Reality
The demo videos are real. The results in the demo videos are not representative. What those videos don't show you is the thirty prompts that came before the one that worked, the hours of post-production that followed, or the fact that the person running the tool had enough creative and technical expertise to recognise which output was worth refining and why.
AI image generation, video tools, and copywriting models are extraordinarily capable but they are not autonomous creative directors. They are instruments. And like any instrument, the quality of what they produce is almost entirely determined by the skill, taste, and intent of the person operating them.
Getting a usable, on-brief, production-ready output from an AI tool typically involves an iterative process that would be familiar to anyone who's ever done a traditional creative shoot: rounds of direction, refinement, rejection, and rework. The tools can be faster than a camera crew. They are not fast in any absolute sense. The labour is still there - it's just wearing different clothes.
The prompt is the brief. And like any brief, a bad one produces work that needs to be thrown away. Writing a good prompt requires exactly the same skills as writing good creative direction.
Myth 02
"Anyone can do it."
The Reality
Access to the tool is not the same as the expertise to use it well. AI has democratised some parts of production. it has not democratised judgement. Knowing whether an output is strategically right, visually coherent, culturally appropriate, and worth developing further requires exactly the same expertise it always did. Arguably more, because the volume of output to evaluate has increased dramatically.
The role of the creative director in an AI-augmented studio hasn't disappeared, it's intensified. Where a traditional creative director might have shaped the direction of ten pieces of work in a week, they're now shaping the direction of a hundred. The filtering, the taste-making, the strategic framing of what gets pursued and what gets discarded, that is all still deeply human work, and it is still the most valuable part of the process.
What AI has done is expand the scope of what's possible within a given time and budget. A small creative team can now explore visual territory that previously required a large production budget. Art direction has become genuinely sharper because you can test, iterate, and pressure-test concepts faster than ever before. But someone still has to know what good looks like, what strategically right looks like, and what on-brand looks like. That someone is still the creative director and that expertise still takes years to develop.
→ Prompt engineering is art direction by another name. The quality of the input (the precision, the reference points, the creative instinct) determines everything about the output.
→ Post-production on AI-generated assets is standard, not optional. Compositing, colour grading, retouching, typography integration; the finishing work that makes something feel considered still requires skilled hands.
→ The expanded scope of output AI enables requires more strategic rigour, not less. Because without a strong creative framework, you're just producing more average work, faster.
Myth 03
"AI saves time, so creative work should cost less now."
The Reality
The labour time has shifted, not disappeared. What used to be spent on production is now spent on direction, iteration, curation, and refinement. The total hours to produce excellent work haven't decreased significantly but the ceiling of what those hours can produce has risen substantially. That's a different value proposition, not a cheaper one.
This is one of the most consequential misconceptions in how clients are approaching AI-augmented creative work. The assumption is that if a tool is faster, the bill should be smaller. But creative value was never priced primarily on production time, it was priced on the expertise, thinking, and strategic judgement that shaped creative impact and commercial outcomes. None of that has gotten cheaper. If anything, in a world where everyone has access to the same tools, the expertise to use them well is worth more than it's ever been.
What AI genuinely changes is the ratio of thinking to production. A higher proportion of the time is now spent on strategy, creative direction, and refinement — and less on execution for its own sake. That's actually a better allocation of talent. It means clients get more thinking per dollar. But it doesn't mean fewer dollars.
Myth 04
"AI-generated creative work is legally clean and ethically straightforward."
The Reality
It is neither, currently. The legal landscape around AI-generated content is genuinely unsettled, and the ethical questions are even less resolved. Using AI tools in professional creative work without understanding these issues isn't just risky, it's a failure of the duty of care every studio owes its clients.
Copyright & Legal
Most large AI image models were trained on datasets that included copyrighted work, often scraped without explicit consent from the creators whose work was used. Several major legal cases are currently working through the courts in the US, UK, and Europe, and the outcomes will have significant implications for commercial use of AI-generated assets.
The current position for commercial creative work is genuinely murky. In many jurisdictions, AI-generated work with no human creative authorship may not qualify for copyright protection - meaning the images you generate may not be ownable in the traditional sense. For brand identity and campaign work, this matters enormously.
Our position: we use AI tools that are trained on licensed or consented datasets wherever possible, we're transparent with clients about where AI has been used in their deliverables, and we stay current on the legal landscape as it develops. We don't treat "everyone else is doing it" as a sufficient answer to an unresolved ethical question.
The Ethical Dimension Beyond Copyright
Copyright is the legal question. The ethical questions are broader and, in some ways, more important for the industry to grapple with.
01. Transparency with clients
If AI tools are contributing to deliverables, clients deserve to know because informed consent is a basic professional standard.
02. The livelihoods question.
The creative industry runs on human talent developed over years of practice. How studios use AI (whether to eliminate that talent or to amplify it) is a values question as much as an efficiency question.
03. Cultural homogenisation.
AI models reflect the biases of their training data. Uncritical use produces work that skews toward dominant aesthetic norms, which is the opposite of what differentiated brand work should do.
04 Misrepresentation
Using AI-generated photography that implies real people, real locations, or real events without disclosure is an integrity issue and increasingly, a legal one.


















