What the Tech Jobs Slowdown Tells Public Affairs About Its Next Decade

By Paul Shotton, Advocacy Strategy

For the past two years, the headlines about technology employment have been uniformly grim. Oracle has announced thousands of job cuts. Block is shedding nearly half of its workforce. The so-called “magnificent seven” of American technology have barely grown their payrolls since 2022, and in San Francisco, overall employment is down three per cent. The easy explanation, repeated in boardrooms and op-eds alike, is that artificial intelligence has finally arrived for the coders.

A recent analysis by The Economist takes issue with that story, and the challenge it poses is interesting for anyone working in a professional services field. The magazine’s examination of technology employment across seven large economies finds the data is consistent: hiring has slowed or stalled since late 2022. However, the timing, the magazine argues, is something of a coincidence. Anthropic’s Claude Code, the first AI tool genuinely capable of replacing a software engineer, only became available in February 2025. Most of the slowdown happened before that. Other forces — rising interest rates, a post-pandemic correction, an acceleration of outsourcing to India, and business investment in software cooling sharply in 2023 — look like better explanations for what has actually happened so far.

If that is right, then the tech labour market becomes a useful lens for public affairs. Our sector is not going to be disrupted by AI on the same timetable, for the simple reason that AI is not yet as capable in public affairs work as it is at writing code. That gap gives us time. So the more interesting question is what we do with it, and whether we are reading the signals from tech correctly, or simply inheriting its panic.

Why programming is the leading indicator

If AI were to reach professional maturity in one field before others, it would be software engineering. Code is unusually well suited to large language models: it is structured, it is verifiable because you can run it to see whether it works, and the training data is abundant, public, and annotated. A coding task also tends to have a clean boundary — write this function, fix this bug — which lets an AI system be tested against a clear success criterion.

Public affairs work has almost none of these characteristics. A stakeholder map or a campaign strategy is not verifiable in the same way. The quality of a briefing depends on context, on organisational politics, and on the reader as much as on the text itself. Even the more structured tasks, such as policy monitoring or legislative analysis, rely on judgement about relevance that is difficult to encode. So it is reasonable to expect a lag: whatever AI does to tech jobs will reach public affairs later, and in a more diffuse way.

That lag should not be mistaken for immunity. It is better understood as advance notice. What happens to software engineers over the next few years will give us a fairly clean preview of the forces that will eventually reshape our own sector, and we can watch those forces unfold without having to live through them first.

Three signals worth reading now

The first signal from the tech labour market is that the loudest explanation is rarely the right one. The AI-driven narrative makes for dramatic commentary, but the Economist’s analysis points to a quieter set of causes that matter more in the short term. Monetary policy discouraged firms from investing in expensive software teams. Companies that had overhired during the pandemic quietly corrected. And outsourcing accelerated: American imports of services related to cloud computing and data storage more than doubled between 2021 and 2024. Why pay a Bay Area salary, the argument goes, when you can get the same output from Bangalore for a quarter of the cost?

Public affairs has its own version of each of these dynamics. Consulting budgets are under pressure in many large organisations, and the first line items to be cut are often external advisers. Clients that scaled up public affairs capacity during regulatory peaks — the Green Deal, the post-pandemic recovery funds, the AI Act — are now reviewing whether they need that capacity at steady state. Offshoring, too, is already underway at the edges of our field, with monitoring, transcription, and basic research increasingly carried out by providers in lower-cost markets, sometimes supported by AI tooling. If our sector begins to see a hiring slowdown in the next few years, it would be a mistake to reach straight for AI as the explanation. Interest rates, post-crisis correction, and outsourcing may matter more, at least initially.

The second signal is more encouraging, and it is perhaps the most under-discussed finding in the Economist piece. Tech jobs are not disappearing. They are spreading. The share of people in the United States who describe themselves as software developers has actually ticked up slightly since 2022. The magazine’s analysis of American occupational data shows the number of computer and software workers employed in retail grew by twelve per cent between 2022 and 2025. In real estate, it grew by seventy-five per cent. In construction, by nearly one hundred per cent. The route to a tech career no longer runs only through Google or Meta. It runs through Starbucks, through house-builders, and through banks that want to act more like software companies.

The public affairs equivalent of this diffusion is already visible in adjacent functions. Sustainability teams, investor relations, ESG reporting, compliance, and corporate communications all increasingly require people who understand how policy is made, how stakeholders are mapped, and how narratives are built. These are public affairs skills even when they do not carry a public affairs title. If tech is a guide, then our sector’s centre of gravity may not shrink so much as migrate, out of Brussels consultancies and trade associations, into the corporate functions that now need to speak credibly about regulation, political risk, and societal expectations. That is not a threat to the profession. It is a widening of it.

The third signal, which runs through the Economist piece and through parallel research on AI adoption more broadly, is that maturity matters. Only twenty-eight per cent of firms in the San Francisco metropolitan area use AI regularly in their day-to-day operations. A survey of firms across four economies by Ivan Yotzov of the Bank of England and colleagues finds the impact of AI on employment has so far been “essentially zero.” The tools are available. The workflows have not caught up. The firms that will eventually capture the value are the ones doing the less glamorous work of redesigning how tasks are actually done.

This is the signal that most directly concerns public affairs management. The maturity curve in our sector is currently uneven. Some teams have invested in structured workflows, shared prompt libraries, consistent taxonomies, and clear processes for monitoring and analysis. Most have not. When AI does begin to reach into the core of public affairs work, in two years or five, the advantage will accrue to the teams that built the underlying discipline early, not to those who simply bought the tools. That is true today for software engineering organisations, and it will be true for ours.

What this means for public affairs teams and consultancies

The practical implication of reading tech as a leading indicator is not to panic, but to prepare, and to prepare for the right things. In practice, three priorities follow.

The first is to resist single-cause explanations. When budgets tighten or hiring slows in public affairs in the coming years, attention will naturally flow to AI as the culprit. It may well be part of the story. More often, however, it will be obscuring a simpler one: cost pressure, outsourcing, or demand correcting after a regulatory peak. Separating these forces matters, because each calls for a different response.

The second is to treat skill diffusion as an opportunity rather than a threat. The fields absorbing tech workers today are the ones investing in serious capability rather than superficial adoption. Public affairs teams that want to remain relevant across the corporate perimeter should be thinking about how they help sustainability, investor relations, and compliance colleagues use public affairs logic, and where the boundaries between these functions now sit.

The third is to invest in the workflow layer before the tool layer. The firms capturing AI’s value today are the ones doing process design, not software purchasing. In public affairs, that means codifying monitoring, briefing, stakeholder mapping, and reporting routines into clear workflows, so that when more capable tools arrive they plug into something structured. Without that foundation, better tools simply produce faster versions of the same fragmented work.

A useful question

If the Economist’s analysis is right, then the story most often told about AI and tech employment is not quite the story the data supports. Other forces did much of the heavy lifting. The AI effect is real and growing, but it is not yet the main event. That is useful for public affairs, because it tells us something about our own near future.

The question worth sitting with, in that case, is not whether AI will disrupt public affairs. It almost certainly will, in time. The more useful question is which of the forces already reshaping our sector — cost pressure, outsourcing, diffusion across corporate functions, uneven organisational maturity — will prove to have mattered most by the time the AI effect becomes visible. Some teams will look back on the next three years as the period in which they quietly built the discipline to make AI genuinely useful. Others will look back on it as the period in which they were distracted by the wrong conversation. Which of those groups do you want your team to be in?