AI strategies for public relations practitioners
AI isn’t a revolution in public relations. It’s a moment of reckoning.
When OpenAI launched ChatGPT in 2022, it triggered an existential crisis for knowledge work.
Three years on, the hot takes in corporate communications and public relations swings back and forth between “AI will take our jobs” and “AI will make us dramatically more effective and efficient.”
Both perspectives are wrong or at least premature.
I’ve tried to slow things down in a chapter for the new book AI and Strategic Communication, edited by Yang Cheng and Dejan Verčič, published last week by Wiley. The aim was to unpack the implications, connect theory to practice, and map tools to organisational systems.
Here’s what I learned in the process
AI is already changing how we work (just not in the way we expected)
There’s no question that AI is having an impact. Practitioners are using it for writing, research, editing, admin, and image generation. Yet most organisations are nowhere near implementing it in a structured or strategic way.
The reason? Governance hasn’t caught up. Neither has clarity about what AI is actually good at.
We’ve learned that AI excels at reductive tasks: summarising, categorising, and recognising patterns. But it struggles with nuance, creativity, and cultural context. The concept of the “jagged edge” (the uneven boundary of what AI can and can’t do) is important. And you only discover it by doing the work.
There’s no AI strategy without a people strategy
Much of the industry has rushed into AI adoption without first establishing the necessary underlying skills or frameworks. Tools are being applied to tasks without understanding how they align with workflows or team competencies.
We can’t keep talking about AI’s opportunities without addressing the risks head-on. Hallucination, misinformation, copyright and regulatory uncertainty are all unresolved issues. Critical thinking and ethical decision-making must be built in.
AI shows up as a shiny new tool, but it necessitates a fundamental shift in the nature of work. If you don’t redesign the workflow, from governance and ethics to training and evaluation, you’re not adopting AI, you’re tinkering and creating a load of risk issues for your organisation.
Knowledge management frameworks, technology adoption models and relationship management theory provide structure. But we also need cultural fluency. Resistance to AI often stems not from fear, but from a lack of context, trust, or relevance.
Use cases beat hypotheticals
Theory helps of course. But case studies tell the real story.
One of the most encouraging examples is Hard Numbers, the public relations agency founded by Darryl Sparey and Paul Stollery. I’ve worked with the agency in an advisory role since 2021.
Hard Numbers is utilising AI for transcription, research, and administrative tasks. They’re also honest about how and where AI underperforms, particularly in content generation. That kind of pragmatism is important.
Moving beyond the hype
We’re past the novelty phase. AI is becoming increasingly embedded in the fabric of how we work, from Microsoft Copilot and large language models to third-party tools.
This chapter doesn’t claim to have all the answers. It’s intended as a foundation. A way to connect the dots between theory, tools, practice, and risk and to help practitioners adopt AI with strategic intent.
Since I wrote the chapter, Ben Verinder and I have created a community of practice to develop this thinking and knowledge. Our book AI for Public Relations: A How-To Guide for Implementation and Management is a work in progress. We’re adding and challenging contributors to update chapters as the submission deadline approaches. Kogan Page will publish the book in May 2026.
Further reading
This essay was originally posted on my Substack. The newsletter is read by more than 5,000 communications and public relations practitioners twice a week. We take a slower, critical perspective to distilling news, research and industry developments into actionable briefings to help you at work.
