What AI Actually Replaces in a Marketing Team (and What It Can't)

This article follows on from my AI upskilling journey published in June 2025. For a function-by-function look at how AI is changing specific marketing roles, see my updated B2B and B2C guides.

Jack Dorsey cut Block's workforce nearly in half last week. Over 4,000 people. His explanation was blunt: intelligence tools have changed what it means to run a company, and a smaller team using those tools can do more and do it better.

The market loved it. Block's share price jumped 16% the day of the announcement. Wall Street rewarded the cut. And Dorsey predicted that within the next year, most companies will reach the same conclusion and make similar structural changes.

The same week, the Microsoft AI CEO said he believes all white-collar jobs could be automated within 18 months. LinkedIn disclosed that its own non-brand B2B traffic had dropped by up to 60% as AI search replaced clicks. And a survey of 1,400 marketers found that 91% now use AI daily but only 41% can prove it's delivering ROI — down from 49% the year before.

If you're a marketing leader, a founder building a team, or a marketer wondering what all of this means for your career, the noise is deafening right now. The headlines oscillate between "AI will replace everyone" and "AI is just a tool." Neither is particularly helpful.

So let me try to be more specific. After two decades running and building marketing teams, and the last few years actively integrating AI into how I work, here's what I actually see AI replacing in marketing, what it genuinely can't replace, and what the smart response looks like.

What AI is genuinely replacing

Let's be honest about this rather than pretending nothing is changing. Some categories of marketing work are being automated at speed, and the people whose roles were primarily defined by these tasks are right to feel the ground shifting.

First-draft production. Writing that blog post from scratch. Drafting that email sequence. Creating the first version of ad copy, product descriptions, social posts, or landing page text. AI does this faster, often at a comparable quality level, and it never has a slow Tuesday. The role of "person who produces the first draft" has been fundamentally devalued. This doesn't mean writers are obsolete — more on that below — but the specific task of generating a first draft is no longer a full-time job.

Routine reporting and data assembly. Pulling numbers from multiple platforms, formatting them into a weekly deck, and adding surface-level commentary. This was never the best use of a human brain, and AI handles it well. The analyst who spent 60% of their week building reports and 40% interpreting them should now be spending 90% on interpretation, because the assembly part is essentially automated.

Repetitive campaign operations. Resizing creative for different platforms. Adjusting bid strategies based on standard rules. A/B testing subject lines across dozens of variants. Scheduling posts at optimal times. Routing leads based on scoring criteria. These tasks are structured, repeatable, and rule-governed — exactly the type of work that AI and automation handle well. The campaign manager whose value was primarily in executing these operational tasks is the role most directly affected.

Basic research and competitive monitoring. Tracking competitor website changes. Monitoring review sites. Summarising industry news. Pulling together a market landscape overview. AI agents can do this continuously rather than sporadically, and they don't get bored or miss things.

Template-based design. Social media graphics that follow brand guidelines. Banner ad variants. Email header images. Presentation formatting. AI design tools can produce these at scale, and the quality gap between AI-generated template work and human-produced template work has essentially closed.

None of this is speculative. These shifts are happening now, in marketing teams of all sizes. The Block layoffs may be extreme in scale, but the underlying dynamic — AI absorbing structured, repetitive, production-oriented tasks — is playing out everywhere.

What AI genuinely cannot replace

This is where the conversation gets more interesting and more important. Because while the headlines focus on what's being automated, the work that AI can't do is becoming more valuable precisely because of that automation.

Strategic judgment. Which market should we enter? How should we position against a new competitor? Should we invest in brand or performance this quarter? What's the right pricing strategy? These questions require an understanding of context, trade-offs, organisational dynamics, and risk tolerance that AI doesn't have. AI can provide data to inform these decisions. It can model scenarios. But the judgment call — the decision where reasonable people could disagree — remains fundamentally human. And it's becoming more important, not less, as the speed of AI-driven execution means strategic mistakes get amplified faster.

Creative originality. Not "creative" in the sense of producing copy or graphics — AI does that competently. Creative in the sense of having an idea that nobody else has had. The unexpected angle. The campaign concept that makes people feel something. The brand voice that's so distinctive it becomes a competitive asset. AI is trained on patterns from existing content. It produces variations on what's already been done. It doesn't take creative risks, because risk-taking requires a willingness to be wrong, and AI is optimised to be statistically safe. The human creative who can produce work that surprises — that makes someone stop and think — is more valuable now than at any point in the last decade.

Relationship and trust. The partnership with a key client that survives a difficult quarter because of genuine human connection. The board presentation where the CMO reads the room and adjusts their message on the fly. The negotiation with an agency partner. The mentoring conversation with a junior team member. Marketing is a deeply relational function, and the relational work — building trust, navigating politics, managing stakeholders, earning credibility — is entirely human.

Taste and editorial judgment. AI can produce a competent blog post. It cannot tell you whether that blog post is worth publishing. It can generate twenty subject line variants. It cannot tell you which one has the right tone for this particular audience at this particular moment. Taste — the ability to look at something and say "this is good enough" or "this isn't right" or "this needs to be bolder" — is a form of pattern recognition that humans develop through experience and cultural immersion. It's the editorial layer that separates content that resonates from content that merely exists.

Ethical judgment and brand stewardship. When a crisis hits, what do you say? When a customer complains publicly, how do you respond? When an AI tool produces something that's technically accurate but tonally wrong, who catches it? When a personalisation engine crosses the line from helpful to intrusive, who decides where that line is? These judgment calls require empathy, values, and an understanding of consequences that AI doesn't possess. The brand steward — the person who protects the company's reputation through every decision — is an irreplaceable role.

Making sense of ambiguity. AI excels when the problem is well-defined and the data is clean. Marketing is frequently neither. Interpreting a confusing set of results. Understanding why a campaign that should have worked didn't. Figuring out what a competitor's move actually signals about the market. Reading between the lines of customer feedback. This sense-making work — taking incomplete, contradictory, or ambiguous information and turning it into a clear direction — is deeply human and increasingly valuable.

The "AI-washing" problem with layoffs

Before we move on, something needs to be said about the Block story specifically, because it illustrates a pattern that marketers should understand.

Block tripled its headcount during the pandemic era. Dorsey himself acknowledged on X that he over-hired because he ran two separate company structures rather than one. The company had already laid off 8% of staff in 2023 and another 8% in 2024, well before this latest cut.

Does AI play a role in enabling a leaner operation? Of course. But cutting from 10,000 to 6,000 is also what a correction from pandemic-era overhiring looks like. Sam Altman himself noted in February that there is some "AI washing where people are blaming AI for layoffs that they would otherwise do."

This matters for marketing leaders because the narrative affects hiring decisions, team morale, and investment strategy. If you believe "AI is replacing 40% of all jobs," you'll make very different decisions than if you understand that "AI is replacing certain types of tasks, and companies are also correcting for over-hiring." The second framing is closer to reality and leads to better planning.

What this means for how you build a marketing team

If AI absorbs the production work and humans become more valuable for judgment, strategy, creativity, and relationships, the implication for team design is significant.

Hire fewer producers, more architects. The most valuable marketing hire in 2026 isn't someone who can write ten blog posts a week. It's someone who can design the content system: decide what gets produced, define the brand voice, set quality standards, and ensure that AI-generated output meets the bar. Think of the shift as moving from a team of writers to a team of editors and strategists — with AI handling the first-draft production that the team then shapes, refines, and directs.

Invest in senior judgment earlier. In a pre-AI marketing team, you could hire junior people for execution and bring in senior people for strategy as the team grew. In an AI-augmented team, you need senior judgment from day one, because AI amplifies whatever strategy you point it at — good or bad. A junior team with powerful AI tools and no senior oversight is a recipe for fast, confident, on-brand mistakes at scale.

Value range over specialism. When AI handles the deep execution in each channel, the human team needs people who can work across channels, connect dots, and see the full picture. The T-shaped marketer — deep in one area, broad across many — becomes more valuable than the pure specialist whose entire value was in executing within a single discipline.

Build AI fluency into every role. This isn't about hiring "AI specialists." It's about expecting every marketer on your team to be competent with AI tools, able to evaluate AI outputs critically, and capable of designing workflows that combine AI efficiency with human judgment. AI fluency isn't a separate skill. It's becoming a baseline expectation, like being able to use a spreadsheet.

Don't skip the governance conversation. As more of your marketing execution runs through AI, someone needs to own the question of oversight. What's the approval process for AI-generated content? Who reviews agent outputs? What happens when something goes wrong? This isn't a full-time role in a small team, but it's a responsibility that needs to sit somewhere specific rather than being nobody's job.

What this means for your career as a marketer

If you're reading this as a marketing professional rather than a team builder, here's what I'd take away from the current moment.

The skills that matter most are the ones AI can't replicate. Strategic thinking. Creative judgment. Stakeholder management. The ability to make sense of ambiguous situations. The confidence to have a point of view and defend it. These have always been valuable. They're now essential.

Production skills alone aren't enough any more. If your primary value is "I can write a solid blog post" or "I can build a marketing report" or "I can manage campaign operations" — and you don't also bring strategic thinking, creative judgment, or leadership capability — then your role is genuinely at risk. That's uncomfortable to hear, but it's better to hear it now than to discover it in a restructuring.

Learn to work with AI, not against it. The marketers who thrive will be the ones who use AI to handle the work they shouldn't be spending time on anyway, freeing themselves up for the higher-value work that builds careers. The ones who resist AI out of principle or fear will find themselves doing manually what their colleagues do in minutes, with no advantage to show for it.

Build your reputation around judgment, not output. In a world where AI can produce volume, the scarce resource is the person who knows what's worth producing. Build your career around being that person — the one with taste, with strategic instinct, with the ability to look at a market and see what others miss.

The week that made it real

This last week in February 2026 will probably be remembered as the moment the "AI and jobs" conversation stopped being theoretical for marketing teams. Block made it concrete. The numbers made it visceral. And the market reaction — rewarding the cuts — made it clear that this pressure isn't going away.

But the right response isn't panic. The right response is clarity about what AI actually does well, what it doesn't, and how to build a team and a career around the distinction.

AI replaces production. It doesn't replace judgment. The teams and the careers that are built around that understanding will be the ones that thrive — not because AI isn't powerful, but because it's powerful in ways that make human judgment more important, not less.