AI is exposing the fault lines in creative production, industry leaders caution
Artificial intelligence can produce creative work faster than ever, but organisations risk accelerating existing – and new – problems unless they redesign the processes surrounding it.
That was one of the central messages to come out of ‘The Role of AI in Creative Production,’ a webinar hosted by Williams Lea and Prolific North.
The discussion brought together Mark Kuhillow, Co-founder and CEO of Trimontium ai, David Hamilton, Director of Business Development at Williams Lea, and Amardeep Devadason, Global Head of Marketing and CGO, Creative and Digital, at Williams Lea.
Hosted by Prolific North Commercial Director, Ben Waterhouse, the session examined how AI is being used across creative production and what the next 18 to 24 months could mean for agencies, brands and production teams.
From experimentation to implementation
The speakers noted that adoption has reached an important turning point. While businesses can now create convincing prototypes rapidly using widely available tools, the harder challenge is becoming how to turn those experiments into reliable systems that can operate at scale.
Mark said the proliferation of usable AI tools had given organisations a much clearer understanding of what the technology could achieve.
“But what does that look like when it’s productionised now?” he asked. “How do we take those prototypes that people are creating within most organisations we’re talking to, and how do we pass them through scalability checks? How do we roll them out? How do we talk about internal adoption? How do we talk about governance?”
That shift from experimentation to implementation, he said, requires businesses to look beyond individual tools and consider the whole creative operating model.
Amardeep observed that the growing number of products available could become a “problem of plenty” for creative teams. The starting point should be a clear understanding of where greater speed, accuracy or capacity is required.
Many tools embedded within platforms such as Adobe and Figma already support everyday productivity. More fundamental change will come when businesses use AI to redesign workflows, he said, rather than simply help individuals complete isolated tasks faster.
“We are still in this place where it’s more about individual and team productivity,” Amardeep said. “How do you get a little more efficient? How do you do something fast? How do you bring something from concept to an asset in a reduced amount of time? That’s the larger question.”
Amardeep Devadason, Global Head of Marketing and CGO, Creative and Digital, at Williams Lea
Rethinking the whole creative workflow
David said the current debate often focuses heavily on generative AI at the final stage of production. Generating assets more quickly will deliver limited value if briefing, approvals, compliance and quality assurance remain unchanged. A production team might create thousands of adaptations in days rather than weeks, only to leave clients and agencies struggling to review them.
“If you turbocharge the engine at one end, is the rest of the process ready?” he said. “No, it’s not in my experience.”
The result could be a faster production engine feeding work into an operation that lacks the capacity to manage it. The bottleneck moves from asset creation to checking, approval and distribution.
David said AI could eventually improve other parts of the workflow, including interrogating briefs, tagging assets and checking image licences. Those applications may prove as important as the generative tools currently attracting most of the attention.
The panel said AI should ideally be considered from the briefing and planning stages of a campaign. That becomes more difficult when several agencies, production partners and client teams are involved, particularly if the production specialist only enters the process after the central creative work has been approved.
Amardeep cautioned against introducing AI as a late attempt to make an existing campaign cheaper or faster.
“You’ve got to be very, very specific about what it is that you want to do differently,” he said. “Or where it is that you need assistance.”
This also requires stronger trust between brands and their creative partners. Clients need to understand where AI is being used, how their information is being handled and whether the technology will genuinely improve the work.
Where AI is already adding value
The panel identified asset adaptation as one of the clearest areas of value today. Once a campaign and its core creative idea have been approved, AI can help adapt that work across different formats, territories, channels and languages.
Campaigns that previously required dozens of assets can now demand hundreds or thousands. AI can help production teams manage that volume, provided creative direction and quality control remain firmly in human hands.
Mark described AI as “amazing at getting to a first pass”, helping teams turn ideas into prototypes at unprecedented speed. However, its work can appear more convincing than it really is.
Using an illustrative 80/20 split, he suggested that an output could be largely strong while still containing weaker or incorrect material. The challenge lies in identifying the valuable work and applying specialist expertise to improve it.
Amardeep described AI as a “force multiplier” during creative development, allowing teams to move rapidly from an early sketch to a prototype or digital mood board.
That greater capacity places more responsibility on creative directors. Producing 150 potential ideas or executions is of little use if nobody exercises the judgement needed to identify the strongest.
“Where I see it right now, the human expert has to be at the helm, and sort of orchestrating what’s being done,” Amardeep said.
The others agreed with David saying creative expertise would become more important as the volume of machine-generated work increased.
“Put it in the hands of creative people, and it’s a hugely powerful tool,” he said.
David Hamilton, Director of Business Development at Williams Lea
The risk of generic creative work
The risk of weak, generic material ran throughout the discussion. The panel warned that indiscriminate adoption was already producing content with similar structures, visual styles and language.
Mark offered a striking example from his own business. His team had received a lengthy brief that had clearly been generated using AI. Before it could respond, it first had to dismantle the document to uncover what the client was really asking for.
“We’ve had a number of briefs in this week, and one of them was very clearly an AI-generated brief, probably nine or ten pages deep,” he said.
“We’ve had to sit down as a team and actually try and deconstruct the brief, to rip out what the AI has put into the brief, to try and get to what a human being would have briefed us with in the first place.”
The example illustrated how apparent productivity gains can simply transfer additional work from one organisation to another. A client can produce a substantial brief quickly, but the recipient may then spend more time extracting its purpose.
Amardeep said the answer lies in the culture of creative teams and the standards they apply to their work. “The use of AI has to be governed,” he said. “It has to be ethical. It has to be responsible, and at all points, it has to be used in context and not indiscriminately.”
Data, governance and ‘shadow AI’
The webinar also highlighted data as a crucial and frequently underestimated part of AI adoption. Mark suggested that access to leading large language models would increasingly resemble a utility. Organisations may switch between providers according to performance and cost, making the quality and structure of their own information a more important source of competitive advantage.
Agencies, he said, could find that transition particularly challenging.
“Agencies don’t tend to be that great at process on the whole,” Mark said. “I think it’s a challenge for the industry because we’re asking it to do what it’s never done before. This genuinely is change management.”
Creative businesses may possess years of knowledge about their clients, audiences and successful campaigns, but much of that expertise remains undocumented or held by individual employees. AI cannot use it consistently unless it is captured, organised and made accessible.
David believes businesses with the most extensive data and the willingness to invest time in training models would be best placed to reduce mistakes. That preparation requires more than entering a brief into a commercially available tool and expecting an immediate result.
Governance is also struggling to keep pace with adoption, the panel remarked. Employees are already using public AI tools without formal approval, creating what Mark described as “shadow AI” inside organisations.
He said leaders should examine how their teams are using the technology, evaluate those use cases and build appropriate policies and technical safeguards around them.
The risks become particularly acute when agencies handle confidential client information. Material shared under a non-disclosure agreement could be entered into an external AI system, raising questions about where the information is stored, where it is processed and whether it can be fully deleted.
The absence of a sovereign UK large language model adds another complication. Mark said that even when information was stored securely in the UK, its processing could still involve technology originating in the US or China.
David said procurement teams were becoming increasingly alert to the issue, particularly in professional services and other regulated industries. AI policies are also beginning to enter contracts and pitch processes.
Amardeep recommended that experimentation should initially take place within controlled environments. Proven applications could then pass through legal, ethical and operational checks before being incorporated into production workflows.
The true cost of AI
The panel also questioned whether today’s low-cost access to powerful AI systems would last.
Mark argued that the leading model providers were effectively subsidising adoption by giving customers access to substantial computing power for relatively modest subscription fees.
“All of the major frontier models are loss leading in terms of what we’re paying for versus what we’re getting,” he said. “At some point, we’re going to be charged an appropriate amount for what we’re using, and that definitely will link into energy costs, and that will see a huge downward revision in terms of what people are actually using it for.”
That could place another constraint on indiscriminate production. If prices begin to reflect the computing power and energy being consumed, businesses may have to become more selective about what they generate and where AI delivers measurable value.
What comes next
Looking ahead, the speakers predicted that AI would gradually become embedded throughout creative and marketing operations and organisations may eventually stop discussing it as a distinct category of technology.
David said the most significant efficiency gains would come from AI operating throughout the creative process, including briefing, asset management, licensing, checking and approval.
Mark’s closing advice was to engage with the technology while maintaining a critical eye.
“Embrace change,” he said. “Embrace the tools, but just keep an eye open as to what they’re churning out.”
ABOUT WILLIAMS LEA
Williams Lea is a global provider of technology-enabled business and marketing services and forms part of RRD.
Its creative production operation supports brands and agencies with high-volume marketing assets across multiple channels, formats and markets. It combines specialist expertise, AI-enabled workflows and round-the-clock global delivery.
Williams Lea employs 15,000 people and serves clients in 20 countries across four continents. Its clients include Nasdaq and FTSE-listed businesses, with particular experience across financial services, legal and professional services.
In the UK, it has offices in Leeds and London, alongside employees embedded within client teams across the country. Williams Lea became a Prolific North Partner in June 2026 after sponsoring the Northern Marketing Festival.
