How will AI disrupt media and advertising in the next year?
From AI-generated ad creative to agentic operating systems, agencies and media owners alike have rushed to invest in AI with the expectation that the technology will change the game, both in the day-to-day and in how business terms are agreed.
To mark The Media Leader‘s inaugural AI Week in Focus, the team asked industry leaders to answer a trio of questions: which AI applications are actually proving value to their business today, which are just hype, and how they think working practices are likely to be disrupted in the medium-term.
Contributors include:
- Dan Hagen, global chief data and technology officer at Havas
- Lea Karam, founder and CEO of Mindscope
- Sajeeda Merali, CEO of the Professional Publishers’ Association (PPA)
- Simon Michaelides, director general of Isba
- Jamie Richards, director of Mediasense
- Emily Roberts, head of digital at the Responsible Marketing Advisory
- Kate Rowlinson, CEO of WPP Media UK
- James Townsend, CEO of Stagwell EMEA
- Freddie Turner, MD EMEA, Chalice
What AI application will prove its value most clearly in the next 12 months?
The one that addresses the people problem.
Jamie Richards, Mediasense: “The clearest value will come from coding-native LLMs like Claude Code.
“These tools make it easier to build internal tools, automate workflows, and prototype solutions with smaller teams. That lowers the barrier to creating bespoke capability across marketing, data, and operations. The value is speed and flexibility. Teams can solve specific problems quickly rather than adapting to generic tools, and iterate in real time.
“They also broaden who can build. Individuals, small agencies, startups, and smaller brands can now develop tools that were previously out of reach, increasing the pace of innovation.
“The risk is that building becomes easier than building well. Without proper engineering oversight and QA, organisations may create systems that don’t scale.”
Emily Roberts, Responsible Marketing Advisory: “Agentic media buying end to end.
“Protocols like AdCP are reaching the point where campaigns that took weeks to build, quality assurance and launch will happen in days, with optimisation beginning earlier and less budget lost to errors that manual processes routinely miss. The value will be visible and measurable. This is no longer a proof of concept and will be growing exponentially in the next 12 months.”
Dan Hagen, Havas: “The one that addresses the people problem. For AI to prove its worth, we need to change the way people work — right through the system — agency, client, partners. I think AI transformation is more of a change management challenge than a purely technical one.”
James Townsend, Stagwell: “AI will begin to prove its value most clearly through applied, customer-facing use cases, with voice assistants leading the way as scalable, always-on interfaces for engagement and commerce.
“AI voice-enabled assistants will move from pilot to production, with The Marketing Cloud’s newvoices.ai and ‘Lou’ for HarrisQuest signalling real traction. Their ability to combine personalisation, utility, and natural interaction will drive measurable ROI in the near term.”
Lea Karam, Mindscope: “Operational decision-making — supply chains, risk modelling, resource allocation. High data density, clear success metrics, low tolerance for ambiguity. This is where AI earns its keep — not in the messy, human work of meaning-making.”
Kate Rowlinson, WPP Media: “Agentic AI powered by unique and differentiated data is the big story — in the future of advertising, the future of work and the future of consumer buying.
“This isn’t a simple win — building good agents is complex, requiring unique insight and proprietary data sets. One of WPP’s Super Agents, the ‘Analogies Agent’, combs through rich historical data to find if problems have been solved effectively in the past, creating a shortcut to inspiration.
“Alongside Super Agents, we are building agents daily, across the business, to solve a whole plethora of client challenges and learn faster. As our recent ‘Advertising in 2030’ report found, marketers predict that most customer interactions with brads will soon be handled agent to agent.”
Sajeeda Merali, PPA: “Those that improve productivity and streamline day-to-day publishing workflows.
“AI can free up time for what matters most – creating high-quality, premium journalism for the engaged communities our publishers serve. In that sense, outputs may become more human, not less.
“Alongside this, AI’s ability to use the extensive audience data our publishers hold, to generate genuinely useful insights that will help publishers and their advertising partners better understand behaviour and reach the right people.”
Simon Michaelides, Isba: “The speed with which LLMs are being developed means it’s almost impossible to answer this. At ISBA we’re keeping an eye on nascent bespoke, walled garden agentic platforms. These are purpose built for marketers and powered by LLMs. The big holdcos as well as boutique agencies like Boldspace are all building interesting offerings in this space.”
Freddie Turner, Chalice: “Sell-side decisioning. The ability for buyers to influence and control decisions at the supply layer has always made sense in theory — the technology just wasn’t there to facilitate it. That’s changed. Containerisation and AI modelling now mean you can deploy campaign-specific logic at the edge, close to the auction, in real time.”
What use case is most at risk of being exposed as nothing but hype?
Efficiency isn’t intelligence. Brands that confuse the two will move faster in the wrong direction.
James Townsend, Stagwell: “The narrative around fully autonomous ad creation is likely to be tested.
“The idea that AI alone can deliver effective advertising will be exposed as overstated. Human judgment remains critical across both creative and media to ensure relevance, differentiation, and effectiveness.”
Lea Karam, Mindscope: “Fully automated media optimisation. If the systems chasing your metrics — impressions, clicks, conversions — are also influencing how they’re generated, the feedback loop starts to distort. You’re optimising against a mirror.
“Efficiency isn’t intelligence. Brands that confuse the two will move faster in the wrong direction.”
Sajeeda Merali, PPA: “The most overhyped use case has to be the idea that AI can actually replace quality journalism!
“While it can generate a lot of content, it cannot replicate publishers’ authority, originality, or deep knowledge about their audiences, qualities that underpin consumer trust and will be increasingly valued in the long-term.”
Dan Hagen, Havas: “The point of activation — delivering and optimising the placements, audiences, content — is such a complex ecosystem.
“The fully automated end to end delivery machine just isn’t there yet — certainly not at scale. Individual delivery into one channel/platform sure, but not across the full comms landscape.”
Kate Rowlinson, WPP Media: “The idea that AI replaces human ingenuity and imagination. New AI tech alone can’t transform the way the industry works. We have to complement it with human creativity to come up with winning ideas and strategies, alongside considered process and talent design to ensure we build teams and new ways of working that can thrive in an AI enabled world.”
Freddie Turner, Chalice: “Any AI business where the technology can’t be explained and the cost is buried in a bundle.
“Our industry is full of incredible technology but also hype and jargon. 2026 will be the year AI should be interrogated — can you actually explain what your AI is doing, who it’s benefitting and how much it costs? If not, it’s likely just hype.”
Jamie Richards, Mediasense: “Fully autonomous marketing, particularly where the models are owned by platforms rather than the brand.
“Solutions like Meta Advantage+ and Google Performance Max can be effective execution tools, but the idea that they can replace human decision-making is overstated. They optimise what they can observe, which is only part of what creates growth. They cannot fully capture long-term brand effects, creative quality beyond short-term signals, or external factors such as pricing, distribution, and competition.
“Advertisers should be clear on three things: what signals these systems can see, what they can’t, and whose interests they are optimising for. Strong fundamentals, independent measurement, and rigorous experimentation still matter, because they address what the platforms can’t see or don’t prioritise.”
Simon Michaelides, Isba: “There have been several blind tests where pure AI creative outputs have been pitted against outputs created by a combination of AI and humans. It’s those created by a combination that wins when judged by a real audience, so the prediction that AI will completely replace humans in the creative process is one that I don’t believe will materialise.”
Emily Roberts, Responsible Marketing Advisory: “The end-to-end AI marketing capabilities built by agencies.
“Agencies have moved fast, and the platforms are real, but there is a growing gap between what has been built and what clients are actually expecting. CMOs were promised cost efficiencies and speed to market, and the honest answer from most agencies right now is that it hasn’t happened yet. Impressive AI infrastructure is one thing. Reduced fees and tangible efficiencies that clients can feel in their day-to-day are another.
“Closing that gap is where the pressure will build over the next 12 months.”
World Media Group members on how AI will reshape the media industry in 2026
What working practice is most likely to be disrupted?
AI is reducing the time required to produce high-quality work across analysis, reporting, and creative. That breaks the link between hours worked and value delivered.
Freddie Turner, Chalice: “Optimising to media metrics. CPMs, CTRs, viewability scores — these became the measurement of programmatic because they were trackable, not necessarily because they reflected what businesses were actually trying to achieve.
“AI that is built from the ground up around business outcomes rather than media proxies changes that fundamentally. AI that optimises to and delivers business objectives should become the standard.”
James Townsend, Stagwell: “Search will see the most significant disruption — not as a decline, but as a shift. Discovery is moving beyond traditional engines into social platforms, AI assistants, and decision-making agents that shape choices before a results page appears.
“The risk for brands is invisibility within these AI-mediated moments of intent and influence. Frameworks like Stagwell Search+ reflect this shift, enabling brands to track presence across AI systems and actively optimise for visibility, sentiment, and outcomes.”
Jamie Richards, Mediasense: “Charging for time rather than value.
“AI is reducing the time required to produce high-quality work across analysis, reporting, and creative. That breaks the link between hours worked and value delivered. If something that took five days now takes five hours, the question becomes: what is the client actually paying for? This creates pressure on time-based models and accelerates a shift towards pricing based on outputs, outcomes, or ongoing advisory value. It also increases the value of judgment. As production becomes cheaper, value shifts to the people that can interpret, challenge, and decide what to do next.”
Lea Karam, Mindscope: “Annual planning cycles.
“Static, calendar-driven strategies are already obsolete — AI enables real-time iteration and dynamic decision-making. But speed without judgement is noise. The advantage will come from automating downstream and moving upstream, where human intelligence defines direction.”
Simon Michaelides, Isba: “The process of gathering information via desk research and basic analysis is something that AI can do more quickly and often better than humans. This is resulting in a reduction in these junior roles at both brand and agency organisations.
“This is something we need to be mindful of when looking at how we equip the talent of the future.”
Sajeeda Merali, PPA: “The biggest shift will be in how audiences discover content, as AI overviews and LLM-driven experiences reshape the journey.
Dan Hagen, Havas: “The strategic process is ripe for disruption. Collating information from many sources, spotting trends, synthesising information, and developing potential routes forward are all great AI use cases. Figuring out the correct application of human and machine in this mix is key — to raise mediocre machine thinking to human brilliance.”
Emily Roberts, Responsible Marketing Advisory: “The billable hours model. Brands will increasingly question paying agency teams for time spent on tasks that AI can complete in minutes. The premium for execution and speed to launch shrinks once those things become table stakes.
“Agencies that adapt and rebuild around strategy, creative thinking and technology investment are well placed for what comes next. Those that don’t risk finding their model under real pressure.”
Kate Rowlinson, WPP Media: “In our industry, it’s probably easier to try and work out what will not be disrupted — as essentially every person, every role, every business needs to apply AI in a disruptive way to remain competitive.”