They Are Birthing Gods in Shadows

Posted by Mike O'Brien on February 24, 2026 in Ai in Marketing, Postcards from the digital frontier | Leave a comment

And some of those gods are coming for your marketing job

I have spent more than four decades in marketing. I started as a copywriter, became a Creative Director, co-founded an agency, trained over 50,000 marketing professionals across the world, and I now teach postgraduate and MBA students at the University of Law Business School.

I have watched this industry absorb wave after wave of disruption – desktop publishing, the internet, search, social, programmatic, mobile. Each time, the same chorus: jobs will disappear. Each time, the same partial truth: some did, others transformed, new ones emerged.

This time, I am less certain that the pattern holds.

What is happening inside the laboratories of OpenAI, Google DeepMind, Anthropic and xAI is categorically different. They are not building better tools. They are building systems that learn to think. And those systems are being pointed, with growing precision, at the cognitive work that marketing has always assumed was safe.

The phrase I keep returning to is… they are birthing gods in shadows.

It is an unsettling image. It is meant to be. Because the marketing profession is sleepwalking into a reckoning that deserves considerably more attention than it is currently getting. I just listened to a Campaign Live podcast where they reflected on leadership conversations around AI’s impact on jobs and, I kid you not, what leaders wear to a pitch. What next? Sleepwalking into oblivion in matching pyjamas?

What is actually being automated right now

Let me be specific, because vagueness is the enemy of useful thinking.

The following marketing functions are not theoretically at risk. They are actively being automated, partially or substantially, today. Not roles, but tasks within roles for now.

Content creation. Generative AI can now produce blog posts, social copy, product descriptions, email sequences, ad variants and long-form articles at a quality level that is, in many cases, indistinguishable from junior to mid-level human output. This is not potential. It is happening in marketing departments and agencies right now, often without formal policy or governing oversight.

Search engine optimisation. AI tools can audit sites, identify keyword gaps, generate optimised content briefs, write and test meta content, and adapt copy in response to algorithm changes – faster, more cheaply, and with greater consistency than traditional SEO teams.

Campaign optimisation. Platforms including Google and Meta have built AI deeply into their performance engines. Smart bidding, automated creative testing, audience expansion, and budget allocation are increasingly removed from human hands. Many of my students arrive believing they will become media planners. They need to understand that much of what media planners do has already changed shape.

Customer insight. Synthesising survey data, social listening outputs, CRM data and behavioural signals into customer understanding was considered a distinctly human skill requiring judgement and pattern recognition. AI now compresses that process dramatically. The insight report that took a team three weeks can now take three hours.

Performance reporting. The creation of dashboards, the narrative interpretation of data, the building of attribution models – all of these are being automated. The marketing analyst role, which proliferated through the 2010s, is under acute pressure.

Visual and creative production. Image generation, video editing, motion graphics, brand asset adaptation – the creative production layer that sits beneath senior creative direction is being restructured in real time.

The honest summary is this: almost everything that is process-driven, template-driven, or data-driven in marketing is now contestable by AI. That covers an enormous portion of what the industry actually does.

The Impact by Level – Who Is Most Exposed

This is the conversation that marketing education is not having clearly enough, and I include myself in that criticism.

Junior and entry-level roles are facing the most acute immediate risk. Not because AI is replacing human judgement at this level, but because organisations are no longer hiring for the execution tasks that used to develop junior talent. The copywriter who wrote product descriptions. The analyst who cleaned data and built weekly reports. The coordinator who managed content calendars. These roles are either disappearing or contracting. The career ladder that previous generations of marketers climbed is losing its lower rungs. This is a structural workforce problem, not just an individual skills problem, and it will have consequences for the talent pipeline that are not yet fully visible.

Mid-level specialists – SEO managers, content managers, paid media managers, social media managers, email marketing managers – face a different but equally serious challenge. Their functional expertise is not becoming worthless. But the volume of skilled human effort that expertise requires is shrinking. One competent marketing professional augmented by AI can now cover ground that previously required a team. The question is whether organisations will reinvest the efficiency gain into deeper marketing capability or simply take it as cost reduction. Based on what I am observing, many are taking the latter path.

Senior and strategic roles are, for now, more protected – but not immune. The CMO or Marketing Director who understands how to brief, interrogate and govern AI-generated strategy has a strong future. The one who doesn’t is a different case. I have seen senior marketers who privately have no idea how these tools work attempting to position themselves as AI leaders. That gap closes. I have also seen organisations that are using AI to compress strategy cycles in ways that reduce the need for expensive senior marketing counsel. Neither dynamic is comfortable.

The unifying truth across all levels is this: the marketing roles that survive and grow will be those centred on human judgement, ethical oversight, creative leadership, relationship intelligence, and strategic synthesis. These are not soft skills. They are the hardest skills. And our profession has historically undervalued and under-trained them.

The governance gap nobody is talking about

Here is where the picture becomes more troubling than a simple jobs analysis.

In most marketing organisations – agencies, brands, in-house teams – AI adoption is happening bottom-up and without governance. Individual team members are using generative AI tools for briefs, for copy, for research, for client communications. Managers are tacitly aware. Leadership often has no formal policy.

This is not unusual in technology adoption. But it creates risks specific to marketing that deserve naming.

Brand integrity depends on consistent voice, values, and judgement. AI-generated content that is not reviewed by someone who deeply understands what a brand stands for is a brand integrity risk. I have seen it happen. Tone-deaf copy. Factually plausible but strategically wrong messaging. Content that passes a spell-check and fails a brand audit.

Data ethics is another exposure. Marketing AI systems are trained on data and trained to use data. The privacy implications of feeding customer data into third-party AI platforms are, in many organisations, unexamined.

Misinformation is a present danger in marketing as much as in media. AI that hallucinates product claims, regulatory information, or competitive comparisons is not a theoretical risk for a team producing high volumes of AI-assisted content without review.

And there is a deeper structural issue. When marketing capability becomes compressed into AI systems, knowledge walks out of the organisation quietly. The institutional memory that sat in experienced people – how a campaign idea was born, why a brand decision was made, what the customer research actually revealed – begins to erode. Speed replaces wisdom. Volume replaces depth.

I call this the governance gap. Capability is scaling faster than oversight. And in marketing, where what we say and how we say it shapes public perception, trust, and commercial relationships, that gap matters enormously.

What employable looks like from here

I want to be direct with you, because I think the marketing industry deserves directness on this rather than reassurance.

The skills that made marketers valuable in 2015 are not sufficient to make them valuable in 2026. That does not mean those skills are worthless – it means they need to be integrated with new ones.

Here is what I am advising my students, and what I would advise any marketing professional at any level:

Become fluent in AI tools without becoming dependent on them. You need to understand what these systems can and cannot do. You need to be able to brief them well, evaluate their outputs critically, and identify their failure modes. You also need to ensure that your own thinking, writing and judgement remain sharp independently. Professionals who outsource all cognition to AI become difficult to distinguish from the AI itself.

Develop the skills that AI cannot replicate at the level required. I mean specifically: genuine creative originality, ethical reasoning, brand empathy, stakeholder relationship management, strategic synthesis across ambiguous information, and leadership under conditions of uncertainty. These are not soft. They are the ceiling of the profession. Build toward them.

Understand data and systems deeply enough to govern them. You do not need to be a data scientist. You do need to understand what data is being used, what it is being used for, and what it cannot tell you. The marketer who can interrogate an AI system’s outputs with informed scepticism is considerably more valuable than the one who accepts them uncritically.

Build and maintain your professional reputation in public. AI can generate content. It cannot generate your perspective, your credibility, or your relationships. Thought leadership built on genuine expertise and visible over time is a form of professional capital that AI cannot replicate and cannot replace.

Commit to continuous learning as a non-negotiable. This is not a skills gap you close once and move on. The landscape is shifting continuously. The marketing professional who does not build structured learning into their practice – not occasional CPD, but regular, serious engagement with how the field is changing – will find themselves managing obsolescence rather than opportunity.

The leadership question

For senior marketers – CMOs, Marketing Directors, agency heads, consultancy partners – the question is not whether AI transforms the profession. It does. The question is what kind of leaders we intend to be during that transformation.

I believe three principles should govern how marketing leaders approach this moment.

The first is governance before scale. Before expanding AI capability across a team or organisation, build the ethical and operational framework within which it will operate. Define what AI-generated content requires human review. Define what data can and cannot be fed into which systems. Define accountability. This is not caution for its own sake. It is leadership.

The second is contribution beyond efficiency. The temptation when AI reduces the cost of content production is to produce more content. Resist it. The question is not how much marketing you can now create – it is how much value that marketing creates. Measure contribution, not volume. Reward thinking, not throughput.

The third is epistemic discipline. In a market flooded with AI-generated content, AI-inflated claims, and AI-assisted hype, the ability to distinguish signal from noise becomes a competitive advantage. Marketing leaders who develop and model this discipline – who ask harder questions, demand better evidence, and maintain genuine scepticism – will make better decisions and build stronger organisations.

Back to the shadows

I began with a phrase that I want to return to now, directly.

They are birthing gods in shadows. In laboratory conditions. Under investor pressure. Inside a competitive dynamic that makes slowing down feel existentially dangerous.

Those gods – the AI systems that are becoming more capable by the month – are not arriving at marketing from outside. They are already inside the briefing room. Already in the content management system. Already in the campaign dashboard.

The marketing profession has always been, at its best, a discipline of human insight applied to human behaviour. The best work I have seen in my forty-plus years has always come from people who understood something true about other people and found a way to say it.

That capacity – to understand, to empathise, to judge, to lead – is not yet automatable. But the conditions in which it can be developed and exercised are being quietly eroded by a profession that is scaling speed faster than it is protecting wisdom.

The gods being birthed in those shadows will be as powerful as we allow them to be. How marketing organises itself, reskills itself, governs itself and leads itself in the next five years will determine whether those gods serve the profession or simply replace it. That is not a technical question. It is a leadership one.