Why your marketing talent needs to be redefined for the AI era

Last November, three Stanford economists published a report on the employment effects of AI. Early-career workers (ages 22 to 25) in the most AI-exposed occupations experienced a 13% relative employment decline since late 2022. The replacement is already in the talent data.

The questions worth asking now: which marketing work stays human? What new skills does the AI era reward? And how do you build them inside a team that’s being thinned from the bottom? Let’s unpack three shifts.

Shift 1. Strategic workforce planning is being shaped by AI needs

C-suite discussions now centre on a fundamental question: which marketing work do humans design, manage, oversee, and decide on, and which work gets handed over to AI? What the hiring data shows:

• AI Search Specialist: 40% of organisations hiring (Jasper).
• Marketing Analytics & Ops listings: 83.9% mention AI (CXL).
• Executive-level marketing roles: 60% mention AI; entry-level: 29% (CXL).
• Emerging titles: AI Marketing Specialist, AI SEO Strategist, Head of AI in Marketing.

Shift 2. AI talent strategy is being rewritten

Career paths inside marketing functions are being redrawn around AI ownership. Performance management is starting to integrate AI use as a competency, and trust in the technology now sits inside the regular team-survey loop.

Shift 3. AI capability development is about judgment, not tools

The AI capabilities being rewarded now sit far from tool competency. The AMA’s 2025 Marketers Skills Report names communication, critical thinking, adaptability, collaboration, and creativity as the top five essential human skills. I have mapped each into AI-era work:

• Critical thinking: know what to ignore from AI output.
• Communication: turn AI insights into decisions.
• Collaboration: work effectively with AI agents and human teams.
• Adaptability: own the consequence when the model is wrong.
• Creativity: direct the AI toward something it would not produce on its own.
• AI governance judgment (one I would add): applying brand, legal, and ethical guardrails inside live workflows — brand compliance, data privacy, copyright, bias mitigation.

The gap is the opportunity

Only 2.66% identify as AI experts. 68% have zero formal AI training. 72% of employers can’t find AI talent. On the other side: employees with advanced AI skills earn 56% more, and the industries most exposed to AI show 27% revenue-per-employee growth versus 9% for the least exposed. Training programmes alone will not close it. Here is a four-week plan.

• Week 1 — Audit the work, not the headcount. For every role, list the three tasks the person spent the most hours on last quarter. Mark each: pre-AI (cannot be done by AI), with-AI (faster with AI in the loop), or post-AI (the AI does it, a human signs off). The with-AI and post-AI columns are your workforce-planning starting point.
• Week 2 — Name the two AI-specific roles you do not yet hire for. The most critical for your team in the next 12 months. Decide whether you build them from inside or hire them in.
• Week 3 — Redefine the judgment layer for every retained role. For each role that stays, write one sentence: the judgments this role makes that the AI cannot. The work shifts from “doing the task” to “deciding which tasks are worth doing and which AI outputs are worth keeping”.
• Week 4 — Wire the operating model so judgment accelerates. Update the job specs. Update the dashboard to measure the judgment humans make, not the number of trainings completed — human-edit rate on AI assets, ROI on AI-assisted campaigns, speed from brief to launch.

The replacement has started. The skills the market rewards have already changed. The teams that win the next two years will audit the work, name the new roles, redefine the judgment layer, and rewire the operating model before it is too late.

← All insights