The Institute of Practitioners in Advertising (IPA) published its 2025 Agency Census last week, reporting on staffing across its member agencies, which manage more than 85% of the UK’s £22bn annual ad spend.
While the data was worth reading, most industry reactions focused on the single narrative that AI is causing a reduction in hiring activity across creative agencies. While it may make for headline-grabbing news, it’s not actually the full story.
The Headline Numbers
Employment across IPA member agencies fell 6.8% year on year, to 24,963. In creative and non-media agencies, the drop was 14.3%, more than 2,000 roles in 12 months. Media agencies grew by 2.4%. The infographic below has the key figures.
The Convenient Narrative
The Census found that 88% of agencies reported that AI is impacting their workflows, but only 8% reduced headcount due to AI in the past 12 mo
That’s quite a significant gap. One describes a change in tools. The other describes a change in hiring. Treating them as the same thing makes for a good headline and a bad diagnosis.
Pull up the five-year IPA data, and the picture changes. IPA member agencies grew by 19% in 2022 during a post-lockdown hiring spree. Current headcount is broadly back to pre-pandemic levels. On a longer timeline, the 2025 reduction seems more like a correction than an exodus.
Then there’s what the AI narrative ignores. The Omnicom-IPG merger has led to thousands of worldwide layoffs. WPP cut around 4,000 roles in 2025 and is folding Ogilvy, VML, and AKQA into a single creative brand. These are financial decisions that have nothing to do with generative AI.
If AI is replacing creative work at scale, where’s the evidence? Which campaign was replaced? Which P&L shows human cost permanently replaced by machine output? AI is changing how agencies work. But a single-year headcount dip during the largest network consolidation in decades is not evidence of that.
The Hidden Problem In The Data
The key question here isn’t what impacted last year’s numbers; it’s what happens next. The Census data on junior talent, graduate intake, and vacancies all point in the same direction: the entry points into the industry are narrowing.
The capability requirements are also changing. AI in creative production doesn’t remove the requirement for talent; it changes what skills they need to be successful in their roles.
Senior and C-level talent hone their skills and capabilities through years of experience working in production-led environments. If the entry-level points close while the capability bar rises, that pipeline comes to a rapid stop.
The agencies reporting the biggest impact from AI are, at the same time, reducing graduate intake and hiring most aggressively. Reduced hiring and rising capability demands don’t cancel each other out; they simply widen the gap.
So, Why Is Talent Choosing To Exit?
Almost 60% of departures in the Census were voluntary, not via redundancy. These were choices made by people, not for them. Some of it is cyclical. Some of it is post-pandemic restlessness. But for early career professionals, AI tools are now doing work that used to be the way in. The career ladder hasn’t disappeared, but the first few rungs are less clear than they used to be.
Agencies are reducing entry-level hiring at the same time. The industry is telling early career talent there’s less room for them, whether that’s the intention or not.
There’s a real difference between agencies using AI to cut costs and agencies building teams that combine creative judgement with AI production capability. The first is a margin decision. The second is a structural one. The first saves money this quarter. The second determines who can compete in two years.
Three things at once
The Census numbers aren’t telling one story. They’re telling three. Holding company restructuring is a market correction driven by financial factors that predate AI. This affects network agencies directly and filters through the wider market.
The talent pipeline contraction is a medium-term risk that most of the industry isn’t treating as one. The experienced creative production professionals that every company will need in five years are being developed now, or they’re not. There’s no way to catch up on that later.
The AI capability question is real but still early. 24% of agencies expect to cut roles due to AI this year, triple the 8% that did so in 2025. Whether that intent becomes action depends on whether agencies can work out what AI-enabled production looks like in practice, not in pitch decks.
From hundreds of conversations with creative production companies over the past year, the businesses getting this right aren’t leading with headcount reduction. They’re building capability, putting people alongside the technology, and working out what the combination can produce. Most of them are doing it quietly.
The agencies that make the right investment in talent now will define what happens next. The ones who treat AI as a reason to stop investing in people will find the cost is much higher than they anticipate further down the line.
