The future of jobs in PR: will we get the third technology shift wrong too?

Latest figures from the Department for Work and Pensions shows nearly a million young people (aged 16-24) in the UK are not in education, employment or training (Neet) and almost three in five are economically inactive.

This is not a case of simply workshy generation, 84% of those who are Neet said they want to work, study or train. Quite simply, the bottom rungs of the career ladder have been kicked away.

This is a phenomenon that the PR industry is not immune from.

As we confront the third major technological shift of the last three decades, the impact on jobs is a question that has preoccupied me for some time and featured prominently in my discussions with industry peers.

There is no settled view in the public relations industry about what AI will do to jobs. Anyone offering you one is probably trying to sell you something.

Here's mine.

This is no longer a theoretical issue

The reductions are real. Three account executives on media monitoring is now one tool. The two-person intranet team is a fraction of that effort.

The government has flagged "public relations professional" among the 20 occupations most exposed to large language models (LLMs). Early-career employment in many of the sectors spotlighted is already in relative decline.

We have been here before. In 1995 we thought the internet was a publishing problem. In 2007 we thought social media was broadcast. The disciplines that emerged, notably search, content marketing and influencer marketing, were largely built by people who weren't us.

I'm hoping that we don't get this one wrong as well.

The source of optimism

In the agencies and in-house teams getting this right, routine work is disappearing. The classic pyramid is becoming a diamond. Smart firms are building roles around insight, outcome and risk management rather than billable hours. If you are still selling time rather than value, AI will break your model.

In-house is different. You can't win another client when efficiency frees up capacity, but internal demand for sophisticated communication skills is rising. As AI absorbs the routine, it makes room for the work corporate teams have always needed: employee engagement, stakeholder mapping, and proactive issues management.

If you want to put numbers on it. Near-term, perhaps 15 to 20 per cent fewer entry positions over one to two years. Three to five years out, a net increase as scope expands. Beyond that, growth in roles we can't yet name, the way digital did after 2000.

Where the caution comes from

The pessimistic case is a robust argument. Martin Ford has been making this case for a decade, most fully in The Rise of the Robots. AI is coming for the professional classes, he says.

The pattern up to now: each wave hits one tier of the workforce, and the tier above absorbs the displaced. This time, Ford argues, there is no tier above. The advisory work that absorbed previous shifts is itself the target.

If he is right, the 1995 and 2007 analogies stop working. The optimistic case rests on a pattern Ford says is breaking. I don't fully buy it, but it’s not easy to dismiss either.

Carl Benedikt Frey and Michael Osborne's hourglass works similarly: strategists at the top, functional roles at the bottom, middle hollowed out. The mentorship model of tacit learning in public relations assumes a middle.

The gender story makes it worse. Women are overrepresented at junior levels and underrepresented in leadership, and are less likely to adopt AI than men. If the bottom of the ladder thins out, the missing women problem also deepens.

There is a quieter cost too. Practitioners report rising anxiety about job security and identity loss. The promise that AI will handle functional roles is unhelpful if that’s the basis of your job.

Which scenario wins, I can’t tell you. What I do know is that we get to decide. We still have human agency. The technology isn't doing the choosing.

Don’t cut off your talent pipeline

Cutting entry-level intake is organisational self-harm. The response I keep hearing is wrong. Freeze apprenticeship and graduate schemes. Let AI cover the production work. Run leaner.

If you do that and in five years you have nobody left who understands how the systems work, why they fail, and when to override them. You cannot run an advisory profession without a pipeline.

The firms that thrive will invest differently. AI-augmented apprenticeships. Judgement taught alongside the tools. Entry-level work redesigned around what AI is genuinely poor at: cultural insight, authenticity checks, creative ideation, and community management.

There's a retention problem running the other way. Junior and mid-career practitioners are walking out of organisations they consider too far behind the curve.

Leadership has been asking the wrong question. Get capability right and headcount looks after itself.

But if Ford is right, none of this is enough. That's the argument I keep losing sleep over.

Stephen Waddington is co-editor of AI for Public Relations: A How-To Guide for Implementation and Management with Ben Verinder, a professional adviser at Wadds Inc, a PhD researcher at Leeds Business School, and a former President of the CIPR.

• Stephen will be picking up these arguments at the AI for PR Conference on 18 June. He'd welcome your perspectives and debate.