For a good while, the loudest conversation around AI in the creative industries was also the least imaginative. Could it make things faster? Cheaper? More abundant? Could it produce a hundred lines of copy, twelve mood boards, a deck summary and a suspiciously polished strategy statement before somebody had even located the meeting room biscuits? Yes, clearly. But that was never the most interesting question.
The one worth asking now is this: how can creative businesses use AI to shape strategy and insight in a way that actually improves decisions, sharpens thinking and leads to better work? That’s where AI starts to matter beyond novelty. Adoption is already widespread. Stanford’s 2025 AI Index found that 78% of organisations reported using AI in 2024, up from 55% the year before, while the share using generative AI in at least one business function more than doubled to 71%.
That shift matters because creative businesses are under pressure from both sides. Clients want sharper strategic thinking and quicker delivery, ideally from the same budget and preferably by Thursday. Teams are expected to understand audience behaviour, market shifts, brand differentiation, channel performance, culture, technology and commercial reality all at once, which is a lovely ambition until somebody has to actually do it. AI, used properly, can help.
But the businesses seeing meaningful value aren’t the ones simply using it to speed up outputs. McKinsey’s latest State of AI reporting points instead to workflow redesign, governance and leadership ownership as the factors that separate shallow experimentation from real business impact.
That distinction is crucial for creative companies. If AI is treated purely as a machine for making more stuff, the result is usually modest efficiency gains plus a fresh assortment of quality-control problems. If it’s used upstream, to synthesise research, test assumptions, map categories, cluster customer language, explore scenarios and sharpen hypotheses, then it becomes something more useful: a way to improve the thinking before the campaign, brand platform, content plan or pitch deck exists.
The OECD’s 2025 review of experimental evidence makes much the same point. Generative AI can support productivity, innovation and entrepreneurship, but the gains depend on how organisations adapt processes and combine the technology with human expertise and trust.
That’s the real opportunity for creative businesses. Not an infinite machine for average content. A better system for turning messy information into clearer strategic direction.
AI is most useful before the work looks finished
Anton Seleznov
A lot of early AI usage in creative teams was entirely understandable. First drafts. Variations. Summaries. Image generation. Production shortcuts. Admin reduction. Nobody should pretend these uses are worthless. They’re often genuinely helpful. But they’re also an easy place to stop, because they’re visible. They create the satisfying illusion of progress. A deck appears faster. A document becomes tidier. Some copy turns up in an unearned tone of confidence. It feels productive.
The problem is that creative businesses aren’t supposed to be rewarded for merely producing more outputs. They’re supposed to be rewarded for making better decisions. Strategy isn’t about the quantity of material a team can generate. It’s about whether the business understands the market well enough to move intelligently within it. Insight isn’t a volume game either. It’s the process of identifying the tension, pattern or behavioural truth that actually changes what a brand says or does.
That’s where AI has far more strategic potential. Instead of asking, “How can this tool help us produce more content?”, the more useful question is, “How can this tool help us understand more before we decide what content or work should exist at all?” That shift sounds subtle. It isn’t. It moves AI from the output layer to the judgment layer.
Recent discussion around AI fluency has pointed to a more useful way of thinking about these tools. The real skill isn’t simply producing prompts or generating outputs on command. It’s defining the business challenge clearly, understanding the strategic end game, and using AI to get to better questions faster. That’s where the technology starts to become valuable rather than merely impressive.
That shift matters because the best use of AI in creative businesses is rarely at the point where the work already looks finished. It tends to be earlier, in the messier stages where teams are trying to frame the problem, interpret signals, compare possibilities and decide what matters. Used properly, AI can help structure thinking, surface first-pass patterns and accelerate the route to usable insight, but it still depends on human scrutiny, strategy and judgment to turn those outputs into something commercially meaningful.
That’s the difference between a gimmick and an advantage. One makes the process look modern. The other makes the thinking better.
Why strategy and insight are the real commercial opportunity
Fuselab Creative
Strategy, stripped of ceremony, is usually the act of turning fragmented evidence into a coherent decision. It sounds grander when written in a keynote font, but that’s mostly what it is. Research comes in from different places. Clients want confidence before certainty exists.
The category is full of recycled language. Stakeholders disagree. Deadlines loom. Somebody wants a bold move, somebody else wants no risk whatsoever, and everybody insists the answer must feel original while still being provably right. Strategy lives in that mess.
Insight is no cleaner. It isn’t some mystical lightning strike reserved for the most tortured planner in the room. More often, it emerges when patterns across behaviour, motivation, context and perception become visible enough to act on. AI can help with both because a great deal of strategic work involves synthesis: reading across large volumes of messy material, comparing signals, spotting repetition, grouping themes, pressure-testing assumptions and pulling meaning out of clutter.
That matters more now because the market is moving quickly. The World Economic Forum’s Future of Jobs Report 2025, based on more than 1,000 employers representing over 14 million workers, found that AI and big data top the list of fastest-growing skills, while skills gaps remain one of the biggest barriers to transformation. This isn’t just a labour-market fact. It’s a signal about how businesses are being forced to operate. They need teams that can interpret complex information faster and make better calls with it.
For creative businesses, that changes the shape of value. The winning agency, studio, consultancy or in-house team isn’t simply the one that can decorate a strategy slide more attractively. It’s the one that can identify what actually matters sooner. Where are the behavioural shifts? Which customer frustrations are persistent rather than anecdotal? Which category codes are exhausted? Which claims are over-owned? Which audience segments are described one way in the brief but talk in an entirely different language in the real world? Those are strategic questions. AI can help teams get through the evidence faster and arrive at stronger hypotheses earlier.
The OECD’s recent work is especially relevant here because it emphasises both task automation and skill enhancement. Generative AI is useful not just when it replaces effort, but when it augments what people can do, especially in research-heavy, synthesis-heavy tasks. That’s precisely where many creative businesses live.
Where AI can genuinely improve strategic thinking
Adriano Parnoffi
The first obvious area is research synthesis. Many creative businesses already sit on more potentially useful information than they know what to do with. Interview transcripts. Workshop notes. Survey verbatims. Search data. CRM notes. Sales-call summaries. Review-site comments. Customer support logs. Competitor websites. Social commentary. Market reports. Presentation decks from previous projects. Usually, these materials exist in large quantities but low usability. They’re available, but not meaningfully structured.
AI can help cluster themes, identify repeated language, group frustrations, compare sentiment across audiences and create a clearer first-pass view of what’s actually showing up in the evidence. That doesn’t mean treating the model like a magical research director. It means using it to do some of the heavy lifting involved in sorting the haystack so humans can more quickly notice which needles matter.
The second area is category and competitor analysis. This is the part of strategic work that’s essential and often deeply tedious. A team needs to review positioning statements, proof points, UX patterns, claims, messaging structures, visual habits, content themes and audience reactions across a category.
Traditionally that involves a lot of manual trawling and a high probability of someone gradually losing the will to live inside a spreadsheet. AI can make this far faster. It can compare recurring language, surface overused claims, identify topic clusters, summarise how competitors talk about themselves and flag where customer feedback contradicts brand messaging. This isn’t glamorous work, but it’s often where sharper positioning starts.
The third area is scenario exploration. Clients rarely need one neat answer that arrives floating down from heaven on a beam of strategic light. More often they need options. What happens if the brand leans premium rather than accessible? What shifts if the audience is defined by behaviour rather than age? What if sustainability is a proof point but not the central proposition? What happens if a challenger tone is adopted in a category where everyone else speaks like a laminated brochure?
AI is useful here because it can quickly generate multiple strategic framings for a human team to interrogate, combine or reject. The point isn’t that the machine gets the answer right. The point is that it gives teams more territory to test before groupthink hardens into certainty.
The fourth area is internal insight. This is arguably the most neglected. Creative businesses often possess rich internal evidence about what they’re good at and where value actually comes from, but they rarely interrogate it properly. Pitch win-loss notes. Project retrospectives. Client feedback. Margin by project type. Repeat business patterns. Brief quality. Reasons for delays. Common objections in proposals.
AI can help identify patterns across these sources, which can then improve positioning, offer design, resource decisions and new-business strategy. A surprising number of agencies still build their story around instincts and mythology when their own data is sitting there trying to be useful.
In each of these cases, AI isn’t replacing strategic judgment. It’s improving the material available to it.
AI should widen the aperture, not flatten the answer
Zara Picken
This is where a lot of AI enthusiasm becomes annoying and occasionally dangerous. Generative systems are extremely good at sounding plausible. They’re very good at filling space with completed patterns. They’re often less good at originality, contextual nuance or strategic bravery. Left unattended, they tend toward average. They predict what might fit, not what’s genuinely sharp, culturally astute or commercially daring.
For creative businesses, that means the risk isn’t only bad information. It’s flattening. If teams use AI carelessly, it can accelerate the production of acceptable nonsense. Not outrageous nonsense. That would almost be refreshing. Acceptable nonsense. The sort that sounds polished enough to slip into a deck and vague enough to survive a quick skim. This is why the best use of AI in strategy is often adversarial rather than obedient.
Ask it to challenge your positioning statement. Ask it what assumptions your brief contains. Ask it where your target audience description drifts into stereotype. Ask it which parts of your proposition sound indistinguishable from category wallpaper. Ask it what evidence would be needed to support the claim you’re making. Ask it to generate the strongest possible critique of your current strategic direction.
That’s where the technology becomes genuinely helpful. Not as a substitute for thought, but as a sparring partner for it.
A growing strand of thinking in the sector is that AI works best as a multiplier rather than a substitute for ideas, direction or expertise. It can help teams move faster through research, synthesis and exploration, but it doesn’t remove the need for critical thinking. If anything, it makes that judgment more important, because a fast answer is only useful when it’s grounded, relevant and strategically sound.
The same principle holds when people debate which parts of the creative industries may become heavily automated. Routine and repeatable tasks may increasingly be absorbed by AI systems, but interpretation, cultural meaning, strategic judgment and emotional intelligence still look far more stubbornly human. That’s why the most interesting question is no longer whether AI can produce something, but whether it can help creative teams understand a problem more clearly before they decide what should be produced in the first place.
Smaller creative businesses can use AI to punch above their weight
Intel
One of the more useful consequences of this shift is that smaller teams now have access to forms of leverage that used to be much harder to come by. A two-person strategy consultancy isn’t suddenly a global intelligence network because it has a chatbot subscription and a brave face. But it can approach a brief with far more structure, pattern analysis and scenario testing than it could’ve managed unaided.
For freelancers, small studios and lean agencies, that matters. Research-heavy strategic work has often been difficult to price properly because it eats time before the output becomes visible. AI doesn’t remove that labour, but it can compress parts of it. It can help analyse transcripts, compare competitor language, organise desk research, summarise market material, generate workshop provocations and identify recurring themes in customer feedback more quickly than manual review alone. That means smaller businesses can spend more of their human effort on interpretation, framing and decision-making rather than drowning in sorting tasks.
The OECD’s review is again useful here because it notes that generative AI can lower barriers for businesses and support entrepreneurship. In practice, that means smaller creative operations can gain capabilities that help them compete more effectively, provided they’re disciplined in how they use them.
Used well, AI can help level parts of the playing field. Not by manufacturing authority, but by helping real expertise move faster.
The hard bit: source quality, provenance and trust
rbl
Of course, none of this means much if the inputs are rubbish.
AI doesn’t rescue weak evidence. It can amplify it beautifully, which isn’t the same thing. If a team feeds in outdated category reviews, poorly designed surveys, unrepresentative comments, vague assumptions and random web findings, the output will still be compromised. It may simply arrive in more elegant paragraphs.
That’s why source quality matters so much. If AI is going to shape strategy and insight, teams need to know where information came from, what it actually represents, what it excludes and how recent it is. Otherwise “AI-powered insight” quickly becomes a more expensive-sounding way of saying “we scraped together some internet soup and let a machine season it.”
There’s also a broader trust question. NIST’s Generative AI Profile, released in 2024 as a companion to its AI Risk Management Framework, exists because generative systems can introduce specific risks around inaccuracy, bias, privacy, security and misleading output. Organisations are expected to manage these risks rather than behave as though the software vendor has handled all the moral thinking for them.
For creative businesses, that translates into some surprisingly unsexy but necessary disciplines. Separate source material from model interpretation. Keep track of what’s first-party evidence and what’s public information. Document major prompts and outputs in higher-stakes projects. Distinguish clearly between observation and inference. Make human review mandatory before strategic outputs are presented as fact. None of this will get anybody onto a panel about the future of creativity. It will, however, prevent a lot of avoidable nonsense.
Data protection and IP aren't side issues for creatives
Grant Barratt
Creative businesses also don’t get to shrug off privacy, data protection and copyright as somebody else’s department problem. These are central issues, especially when AI is being used on real client work, customer material or internal business data.
The UK’s Information Commissioner’s Office is explicit that AI use sits within data protection obligations and offers detailed guidance on applying UK GDPR principles to AI systems. If teams are uploading customer data, interview transcripts, support logs, research notes or sensitive internal information into AI tools, they need to understand what those tools do with the data and whether that usage is appropriate and lawful.
Then there’s copyright, which is no minor subplot for a magazine read by people whose livelihoods depend on original work. In the UK, this debate is intensifying rather than fading. Recent parliamentary scrutiny has warned that the sector faces serious risks from generative AI if copyright protections are weakened. That matters because the creative industries aren’t some niche side hustle the economy occasionally remembers at awards season. They’re a major part of the UK economy and one of the sectors most exposed to how AI is governed, commercialised and contested.
So yes, use AI to improve research, analysis and strategic clarity. But don’t pretend the legal and ethical terrain is settled. It isn’t.
Regulation is starting to shape the environment
Peter Schmidt Group
Even businesses that find regulation boring, which is to say almost all of them, now need at least a basic grasp of the landscape. This doesn’t mean every agency founder now needs to cosplay as a regulatory specialist. It does mean that “we’re just using a few AI tools for ideation and research” is no longer a serious operating model.
Teams need to know which tools they’re using, what data is going into them, what review steps are required and what can’t be delegated to automated systems. McKinsey’s research repeatedly points to leadership ownership and workflow redesign as the things that turn AI usage into business value. That’s another way of saying somebody adult needs to be in charge.
For creative businesses, that ownership question is especially important because the temptation is to let AI adoption happen informally. One planner uses one tool. One account lead uses another. Design experiments with a third. Somebody in strategy quietly starts pasting transcripts into a model at midnight because the workshop is tomorrow and everyone’s tired. Before long, the business has “adopted AI” in the same way a teenager “cleans their room”: technically something has happened, but nobody would call it a system.
Real value comes from deciding what AI is for. Which workflows matter most? Which research tasks benefit from support? Which strategic outputs require human sign-off? Which data is off limits? Which use cases are acceptable and which are lazy? Without those decisions, adoption stays shallow.
A practical framework for using AI to shape strategy and insight
Phenomenon Studio
The most sensible way for creative businesses to approach this isn’t as a revolution, but as a disciplined operating framework.
- First, define the problem properly. Not “use AI to find insights,” which is the kind of brief a machine would write after reading too many LinkedIn posts, but something specific. Why is brand consideration falling? What tensions are emerging in the category? What are customers repeatedly complaining about? Which emotional spaces are already overcrowded? What’s stopping a proposition from feeling distinctive?
- Second, gather evidence deliberately. Pull together first-party information, research material, interview transcripts, search data, competitor messaging, market reporting and customer language. Separate observed evidence from assumptions. Know what’s recent and what’s stale.
- Third, use AI for structuring and synthesis. Cluster themes. Compare language. Identify gaps. Surface contradictions. Summarise repeated concerns. Generate alternate framings. Map possible strategic territories.
- Fourth, apply human interrogation. Which outputs are genuinely useful? Which are generic filler? Which are unsupported? Which feel strategically interesting? Which simply restate the obvious in smoother prose?
- Fifth, run scenario tests. Pressure-test the proposition. Stress the audience definition. Ask what competitors could imitate easily. Ask what evidence is missing. Ask what happens if the market changes or a different stakeholder lens is applied.
- Finally, convert it into action. Better briefs. Better positioning. Better workshop inputs. Better decision-making. Better creative direction.
This is where AI earns its keep. Not by acting as a replacement strategist, but by helping strategists and decision-makers work with a broader, better-structured field of information.
What humans still do better, and probably will for quite some time
Friend + Johnson
Despite the more theatrical claims made online, human strategists, editors, planners, founders and creative leaders still do several things far better than current AI systems.
They understand context in full colour. They detect political sensitivity inside businesses. They notice when the stated problem isn’t the real one. They sense when a brand is hiding behind language that sounds confident but means nothing. They can tell the difference between an interesting observation and an insight that genuinely changes the work. They understand culture as lived reality rather than just text patterns. They know when the bold answer is right and when it’s merely performative.
They also know that strategy is social. It has to be sold, explained, defended and made useful to other humans. A model may help produce options, but it doesn’t lead the client through uncertainty. It doesn’t read the room. It doesn’t recognise that the most analytically correct argument may fail completely if it’s framed in the wrong way for the people making the decision.
That means the future isn’t “AI replaces strategy.” It’s closer to “AI changes what good strategy work looks like.” The routine synthesis becomes faster. The evidential base can become broader. The questions can get sharper sooner. But interpretation, persuasion, taste, contextual judgment and meaningful originality remain stubbornly human territories.
Which is, frankly, a relief.
The real shift for creative businesses
Something Big
The deeper change here isn’t just technological. It’s operational and cultural.
Creative businesses have traditionally described themselves as makers. They make brands, campaigns, content, experiences, products, films, designs, activations and ideas. All true. Increasingly, though, the more resilient ones will also need to think of themselves as intelligence systems. Not in a sinister robot-overlord sense. In the useful sense. They gather signals well. They structure knowledge well. They notice patterns quickly. They interrogate assumptions. They turn evidence into direction before the market has fully moved on.
The wider conversation around AI in the creative industries has increasingly focused on fluency rather than novelty: knowing what the tool is for, how the output will be used and how it connects back to a real business objective. That’s a healthier framing than the usual obsession with whichever model, workflow or trick happens to be fashionable this week. AI becomes useful when it’s tied to outcomes, not when it’s treated as a parlour trick with a software subscription.
That’s the challenge for agencies, studios, in-house teams and independents alike. Not to become “an AI business” in the most tedious possible sense, but to become a more perceptive creative business. One that uses AI where it strengthens diagnosis, research, analysis and scenario planning, and resists it where it encourages generic thinking, careless sourcing or strategic laziness.
Because that’s ultimately how creative businesses can use AI to shape strategy and insight: not by handing over judgment to a machine, and not by stapling a few shiny tools onto yesterday’s workflow, but by using AI to widen research, organise evidence, interrogate assumptions and improve the quality of decisions. Adoption is already here. The difference now lies in how intelligently it’s applied. The businesses that win won’t be the ones producing the most AI-assisted noise. They’ll be the ones using better evidence to make better calls, and turning those better calls into creative work people can actually believe in.