Among all major industries adopting artificial intelligence, marketing stands out as one where AI is no longer confined to experimentation. Instead of living in innovation labs, AI is embedded directly into briefs, production pipelines, approvals, and media optimisation. Insights shared by WPP iQ, drawing on collaboration with Stability AI, offer a practical view into what operational AI looks like when it becomes part of daily agency work.
The emphasis is no longer on novelty, but on whether AI meaningfully reduces friction or simply introduces another layer of tools and complexity.
Turning brand accuracy into a system, not a manual task
For marketing agencies, brand consistency has become something to engineer rather than enforce manually. Off-the-shelf AI models are not trained on a brand’s specific visual language, which often results in outputs that feel generic or off-brand.
To address this, agencies are fine-tuning models on brand-specific datasets so they internalise visual identity, tone, colours, and stylistic rules. Once trained, these models can reproduce brand elements consistently across assets.
WPP’s work with Argos illustrates this approach. After fine-tuning a model for the retailer, the system learned not just character design, but lighting, shadows, and animation details unique to the brand’s 3D style. These are precisely the details that traditionally consume time through re-rendering and multiple approval cycles. When AI outputs start closer to a finished state, creative teams spend less time correcting and more time refining narratives and adapting content for different channels.
Production timelines shrink, but constraints don’t disappear
Traditional 3D animation struggles to support reactive marketing, where cultural moments require rapid turnaround rather than long production cycles. In the Argos case, WPP trained custom models on specific 3D characters, teaching the system proportions, behaviours, and interaction details.
The result was high-quality visual content generated in minutes instead of months.
However, faster generation doesn’t eliminate bottlenecks — it shifts them. As asset creation accelerates, review processes, compliance checks, rights management, and distribution become the new constraints. These challenges existed before, but AI makes them more visible by dramatically reducing production time. Agencies seeking real operational change must redesign workflows around AI, rather than simply adding it as another tool.
The interface problem becomes a strategic issue
WPP and Stability AI also highlight a growing “UI problem”. Creative teams often lose time navigating disconnected and overly complex tools, moving assets manually between systems and relying on workarounds.
The response has been the development of brand-specific AI front ends, with simplified user experiences layered over complex backend workflows. WPP positions its WPP Open platform as a way to encode proprietary knowledge into accessible AI agents that support planning, production, media creation, and sales. Efficiency gains come from cleaner handoffs as work flows from brief to production, activation, and performance feedback.
Self-service AI reshapes the agency-client relationship
AI-powered platforms are increasingly client-facing, allowing brands to self-serve certain marketing functions. This pushes agencies to concentrate on areas that are harder to automate, such as designing brand systems, building and maintaining fine-tuned models, and embedding governance into workflows.
As clients gain more direct access to AI tools, agency value shifts toward architecture, oversight, and strategic control rather than execution alone.
Governance moves into the workflow itself
For AI to be used safely and consistently, governance must live where work happens. Dentsu describes creating secure “walled gardens” where teams can prototype AI-enabled solutions without exposing sensitive data. Successful experiments can then be promoted into production environments.
This approach reduces risk while allowing innovation to scale, turning governance from a static policy into a practical part of daily operations.
Strategy and insight accelerate alongside production
AI’s impact extends beyond creative output. Publicis Sapient describes AI-powered planning systems that compress months of research into minutes by combining large language models with contextual data and structured prompt libraries. Faster insight generation allows agencies to respond more quickly to shifts in culture, platforms, and audience behaviour — and to serve more clients without expanding timelines.
What this means for marketing professionals
Across agencies, job roles are evolving rather than disappearing. Time spent on repetitive drafting, resizing, and versioning declines, while responsibility for brand stewardship increases. New operational roles are emerging, including model trainers, workflow designers, and AI governance leads.
AI delivers the greatest operational impact when agencies rely on customised models, intuitive front ends that reduce friction for both teams and clients, and integrated platforms that connect planning, production, and execution.
While speed and scale are the most visible benefits, the deeper shift is structural. Marketing delivery increasingly resembles a software-enabled supply chain — standardised where possible, flexible where necessary, and measurable at every stage.
