If you run or work in an agency, the honest answer to "will AI take our jobs?" is this: it is taking tasks, and the tasks it takes first are the ones we used to give juniors. That makes the entry-level role the pressure point. The agencies that come out of this well will redesign junior jobs around judgement, not quietly stop hiring graduates.
The quiet part is now being said out loud. At a town hall during its AI Marketing Strategies event, Digiday heard the owner of a small-to-mid-sized performance agency say: "We haven't hired college-level grads in some time, and I don't anticipate that we will be." A day after Digiday published, Marketing Dive reported Gartner's prediction that most high-performing marketing teams will eliminate traditional entry-level jobs by 2030.
Is AI actually taking agency jobs?
Some, yes. The numbers are less apocalyptic than the headlines and more specific. Gartner surveyed 1,303 senior marketing leaders between January and April 2026. According to Marketing Dive, 18% had already eliminated some positions because of AI. Sixteen per cent had created new roles. Nearly a third had reworked existing functions.
Look at which number is biggest. Reworking. Most organisations are not deleting departments; they are changing what the people in them do. The jobs that disappear are the ones that were really a bundle of repeatable tasks with a title stapled on.
The agency owner quoted by Digiday was not anti-graduate. His firm used to hire only entry-level talent, and most of those hires became among its best employees. "It's just a business issue," he said. That is the trap: a decision that is rational for one year's P&L and corrosive for the next decade's talent bench.
Which agency roles are most exposed to AI?
Stop thinking in job titles and start thinking in tasks. Every agency has a list of chores that eat junior hours: timesheets, monthly reports, status meetings, budget tracking, manual resizing, competitor audits. Those are exactly the chores AI and automation now handle well. Judgement, taste and client trust are not on that list.
Here is how the work shifts for a typical junior in a creative, media or social agency.
| Junior task today | What AI now does | What the junior should do instead |
|---|---|---|
| Building the monthly performance report | Pulls the numbers and drafts the commentary | Spot the one change that matters and explain it to the client |
| Competitor audit by screenshot | Collects and groups competitor ads and posts | Decide which competitor move is a threat and which is noise |
| Social listening summary | Clusters conversation into themes | Test whether a theme is a real audience tension or a loud minority |
| Resizing assets for every format | Generates the variants | Check brand fit and catch what looks wrong |
| Status meeting notes and actions | Transcribes the call and lists actions | Chase the decision that is blocking the work |
The right-hand column is harder than the middle one. It is also more interesting, more visible to clients and a much faster way to learn the business.
Why is cutting graduate hiring a false economy?
Today's account executive is 2032's account director. Stop hiring at the bottom and in five years you will be buying experienced people at the top, bidding against holding networks with deeper pockets. For an independent agency, that is a bad trade.
Gartner's advice to CMOs, as reported by Marketing Dive, applies just as well to agency leaders: do not use AI efficiencies as a reason to cut new talent out of the pipeline. Kristina LaRocca-Cerrone, a vice president analyst at Gartner, put it simply: "CMOs should give junior marketers exposure to higher-value work earlier." Gartner's examples are useful ones: evaluating AI outputs, generating insights and linking recommendations to measurable outcomes.
One attendee at the Digiday town hall framed it better than most consultants could: "It seems like a learning and development challenge, not a technology challenge."
Why does AI make capable people feel like frauds?
There is a psychological layer here that most AI roadmaps ignore. Psychologists call it the imposter phenomenon: the persistent sense that you are less competent than people think and about to be found out, whatever the evidence says.
Agencies are prime territory for it. Juniors often prove their worth through speed and stamina on grunt work. Take that work away and many feel they have nothing left to show. Seniors are not immune. One participant at the Digiday event summed up the fear: "AI isn't coming to replace your job. Someone who knows AI and has your experience is coming to replace your job."
People who feel like frauds behave in predictable ways. They over-polish instead of shipping. They don't ask the obvious question in the meeting. They hide uncertainty rather than flag it, which is the last thing you want from someone checking AI output. Authenticity thrives where insecurity ebbs. Give people a clear, valued role and the tools to answer their own questions, and they start offering their real opinions.
How do you redesign a junior agency role for AI?
Start small: one team, one client, one quarter.
- Audit the chores. List every recurring task juniors do in a month and the hours each one takes. Mark the ones AI or automation can now handle.
- Rewrite the job description around questions, not outputs. "Produces the monthly report" becomes "explains what changed for the client, and why".
- Make checking the machine a named skill. Give juniors explicit responsibility for verifying AI output against the source data, and credit them when they catch an error.
- Give every junior a weekly client question. Something real, such as "why did engagement on the competitor's new campaign jump?", answered with evidence and presented to a senior.
- Pair juniors with seniors. One Digiday participant noted that older staff hold "historical knowledge about like just how the sauce gets made". Juniors bring fluency with the tools. Trade both ways.
- Hire for judgement. Gartner suggests assessing candidates on how they work with AI, including their critical thinking. Set a task built on an AI-generated analysis with a deliberate mistake in it, and see who spots it.
What should agency leaders tell their teams now?
The truth, with a plan attached. Telling people nothing will change is not credible. Telling them AI will do everything is not true. One executive at the Digiday event described a sensible sequence: quick wins first, then clarity about the pace of change, then formal training.
The tooling matters too. If the only way into client and category data is a dashboard with forty filters, juniors will keep fetching rather than thinking. Defy MCP lets a team interrogate that data inside Claude and prompt it like a co-worker, with no dashboard to learn, which turns the apprenticeship from "go and get the numbers" into "go and find out why". That is the work that builds strategists, and you can see where it sits in the Defy platform.
The agencies that stop hiring juniors will save a salary line this year. The ones that rebuild the ladder will own the next generation of talent.
