AI: First, they came for the press release…

How did you learn your craft?

I learned mine by writing things that weren't good enough, repeatedly, in front of people who told me so. A press notice rewritten four times. A briefing for a political party spokesman that came back covered in red ink. Maybe also saying something idiotic in a meeting. It may not have been efficient, but it was certainly formative.

That model of professional development is now breaking across the global business landscape. Analysis of 62 million CVs found that AI-adopting firms have cut junior hiring by 9% to 10% within six quarters, while senior hiring was unaffected. AI is not displacing all the work. But it is displacing the entry point.

There is growing noise from the agency side about the future of the pyramid model (clue: it has no future). But what about the (many more) in-house comms roles? And what about the future CCO?

An MIT Media Lab study published last year measured brain activity in three groups writing essays: one using ChatGPT, one using search engines, one using nothing. The AI group produced polished, well-structured work in the fastest time. They also showed measurably weaker brain connectivity, worse recall of what they had written, and a fading sense of ownership over their own words. Some could not quote a single line from an essay they had submitted minutes earlier.

More and more studies are finding a strong negative correlation between frequent AI use and critical thinking, driven by what researchers call cognitive offloading. 

We can all see that this matters. That the production work we are now automating away was never just output, but provided critical learning. But have we really digested quite how much this might matter for our function?

Every comms person believes they write well. Some do. But writing was never really the point. Organising an argument on paper is how we think. If you are not practised at it, you are not thinking enough. And you will certainly not become the editor the new team structures will require, because you cannot edit well what you could never write well. And this refers to everything – not just press releases but annual reports, strategy papers, board papers, not to mention the many emails we write internally every day to interpret, explain, persuade, flatter and cajole.

The counterargument is that junior staff will simply learn different, higher-order skills. Perhaps. But judgement is not downloaded, it is accreted. The law firms taking this seriously are effectively paying juniors to learn, with some allowing first-years hundreds of billable hours for AI experimentation and structured review. I have yet to see a comms function do anything equivalent. Most are heading the other way, quietly not replacing the departing junior because the team ‘copes fine’ with AI.

Follow this through a decade. The production layer compressed. The apprenticeship gone. The craft skills of the mid-level cohort thinner than the generation before. 

So then you have to ask: why would a CEO in the future promote or hire a CCO that didn’t have many years of learnt experience and skills? In fact, why appoint a comms lifer at all?  

Let’s be honest with ourselves here. In the companies I’ve worked for, I’ve come across plenty of very smart, motivated folk, good operators, who have specialised in growing into new roles on their way to senior leadership roles. Maybe originally from finance, from sales, or operations, or marketing. Even management consultancy. 

Right now, very few of these people could do what we do. But as the barriers to doing the ‘practice’ of comms break down, why wouldn’t these people rotate in and through comms roles, supported by small, AI-tooled experts? And once a few have done this, why wouldn’t a CEO appoint one to the top CCO role?  After all they come from the business, know in detail how it is run, have a broader professional hinterland than most of us, and would have the smarts and the training to translate that into value-adding comms functions.

I’ve already seen this happen to other functions, notably including HR.  And the barriers to entry to do what we do are far lower…

The answer cannot be to ban the AI tools, of course, or to pretend the production layer can be preserved as a training ground for its own sake. But perhaps we need to rebuild development deliberately: protected writing time, structured editing under senior scrutiny, rotation through the work AI cannot do, and honest frameworks that define what comms expertise actually is. We need to decide how our people will learn the craft, because the old way of learning it is going, and it is not coming back.

First AI came for the press release. How much more will we let it subsume, and at what cost?

Barney Wyld has held CCO or Corporate Affairs Director roles at companies including Rolls-Royce plc, National Grid plc, Network Rail and Johnson Matthey. He has been a member of the Page Society for seven years and is an active contributor to the Society's work on the evolving role of the CCO. Barney currently uses his expertise to help CCOs review the structure and operating models for their teams.