Forget the hopeful articles suggesting AI will be a mere "co-pilot" for marketers. That narrative is a comforting fiction. For public relations managers, the reality is far more brutal and immediate. By 2026, agentic AI will not just assist, but entirely autonomously execute huge swathes of the traditional PR workflow. The question is no longer if AI can do the job, but how much of it you're willing to let it take.
To pretend otherwise is to prepare for obsolescence. The PR manager of today, buried in coverage reports and press release distribution, is an endangered species. The PR manager of 2026 will be a strategist, a crisis-navigator, and a networker-in-chief – the purely functional tasks having been ceded to silicon.
To understand the shift, we must be ruthlessly honest about the limitations of human execution. A human PR manager, however talented, is a single node operating on a biological clock. They sleep, take holidays, and can only process a finite amount of information. Agentic AI has no such constraints.
A junior PR executive at a London agency might spend their entire morning scanning for client mentions. They use tools like Cision or Meltwater, setting up alerts and manually compiling reports. This process is slow, expensive, and prone to human error. A typical subscription for these platforms can run from £5,000 to £25,000 annually, layered on top of staff salaries.
Agentic AI systems, however, can ingest and analyse the entire global media landscape in real-time, for a fraction of the cost. They don't just spot keywords; they perform sentiment analysis, identify emerging narratives, and even predict the potential trajectory of a story. A human might spot that a Guardian journalist has mentioned a competitor; an AI agent can tell you that this journalist writes about supply chains every third Tuesday, has a negative sentiment towards offshore manufacturing, and is currently engaging with ESG-focused academics on X (formerly Twitter). The AI provides a level of predictive insight a human team could never match.
The boilerplate press release is dead. For years, PRs have churned out formulaic announcements that journalists largely ignore. Writing a standard release for a product launch or a new hire is a task that AI now executes flawlessly. More than that, it can tailor that single announcement into a dozen different formats: a LinkedIn article for the CEO, a series of tweets, a formal release for the wire services, and a blog post for the company website.
UK brands are already leveraging this. Look at the sheer volume and consistency of content from digital-first finance brands like Revolut or Monzo. While they retain skilled comms teams, the underlying engine for much of their day-to-day announcements and market updates is increasingly automated. An agent can draft a release, build a hyper-targeted journalist list from a database like Propel or Prowly based on recent writing (not just a static "tech journalist" list), and schedule personalised distribution. The human role becomes one of final approval and strategic oversight, not manual execution.
Ask a PR manager for a coverage report and you’ll often get a PDF of clippings and a list of Domain Authorities. This is vanity reporting. It tells the board nothing about actual business impact. Did the coverage in The Telegraph actually drive traffic to the website? Did the mention on a podcast lead to sales?
An integrated AI marketing system connects PR activity directly to business outcomes. By tracking UTM parameters, website traffic, and even brand search volume (via tools like Google Trends), AI can draw a direct line between a piece of coverage and a commercial result. It can definitively state that the article in Marketing Week generated 350 visits, 12 demo requests, and influenced £50,000 in the sales pipeline. This is the language of the board, and it’s a language AI speaks more fluently than most human PRs.
The Contrarian Take: AI Lacks a Soul (And a Network)
This is where many AI-sceptics plant their flag. They argue that public relations is fundamentally about human relationships. The long lunch with a key editor, the trusted phone call to kill a negative story, the nuanced understanding of a journalist's ego – these are things a machine cannot replicate. And they are, for now, correct.
However, the contrarian take is not that AI will fail at this. The real take is that 90% of what most PR managers currently call "relationship building" is actually low-value, transactional communication that can and will be automated. The endless "checking in" emails, the generic event invitations, the holiday cards – this is not relationship building. It is administrative fluff.