Internal Side of Public Affairs – 73 AI and Public Affairs Process-Structure

By Alan Hardacre, PhD

Co-Founder Advocacy Academy, Advocacy Strategy

Public Affairs has always used technology. Monitoring platforms, CRM systems, stakeholder databases — none of this is new. What is currently new is the scale and speed at which AI is changing what those tools can do, and more importantly, what your team is expected to do with them. I am sure everyone has heard in their organisation that 2026 is the year of AI.

We are at an inflection point. The question is no longer whether to use AI in Public Affairs. It is whether you are using it in a way that creates value — or just adds noise to an already busy workflow. Let me share what I am seeing at the moment.

The problem with how most teams approach it

Most Public Affairs teams are experimenting with AI. Very few are getting results.

The pattern is familiar: individuals adopt tools reactively — an Enterprise AI license here, and several private licenses there. There are no shared prompts, no common standards, no governance and no team identified use cases. AI knowledge accumulates with individuals and their own tools rather than in the function and their tools. And because there is no baseline, it is almost impossible to measure whether the investment is working – is it more effective? Is it saving time? Is the quality better? In fact what are the metrics we are judging this against?

All of this is not an AI problem. It is a Public Affairs process and structure problem.

The equation that actually matters

Here is the uncomfortable truth: AI tools are not the differentiator here. Anyone can access ChatGPT, Claude, Copilot, or Perplexity. The tools themselves are broadly available and broadly similar. What creates value in Public Affairs now is the combination of three things: experience, intellectual property, and AI.

Experience means knowing what good looks like — how to ask the right questions, how to interrogate an output, how to advise the business on what a policy development actually means for operations. It means seeing and add nuance. IP means the data and frameworks that are uniquely yours: your stakeholder maps, your messaging documents, your issue tracking, your institutional memory. The intellectual property here is all about process-structure and governance. AI is the amplifier — it multiplies the value of experience and IP when applied with structure.

An experienced Public Affairs team with no AI will become slower relative to peers. An AI tool with no experience or IP behind it produces generic output that looks competent but isn't. The teams pulling ahead are those combining all three — structure, expertise, and proprietary data — into workflows that are repeatable and scalable. This is where the real value lies.

Getting the level right

Not every team needs to build custom AI agents and workflows. The maturity journey in AI is genuinely incremental, and jumping too far ahead creates more problems than it solves.

Most teams start at individual use — using a general LLM to draft, summarize, generate content or research. That is a useful beginning. The next step is building shared custom agents, which forces a team conversation about how work should actually be done: what assumptions to bake in, what quality standards to apply, what tone and format stakeholders expect. That conversation is valuable even before the AI does anything.

From there, teams can start looking at workflow automation — not just individual tasks, but the handovers between them. Monitoring alerts that feed into briefings. Stakeholder data that populates engagement plans. Reporting outputs that flow automatically to leadership. The complexity grows, but so does the return.

The point is not to rush to the most sophisticated solution. It is to build capability step by step, starting from where your team is today.

So many use cases

The scope of what AI can genuinely help with in Public Affairs is broader than most teams realize. From extracting data from regulatory documents and mapping stakeholder networks to drafting position papers, running sentiment analysis, and modelling policy scenarios — the use cases span every part of the PA workflow.

That breadth can be paralyzing if you try to tackle it all at once. The more useful frame is to ask: where is my team spending the most time on tasks that are largely mechanical? Start there. Build confidence. Then expand. That is where the leading teams started.

The question behind the question

There is a slide I use with clients that frames the challenge cleanly. Most Public Affairs teams sit at the individual-use level, with occasional experimentation at the team level. Very few have embedded AI into their function's operating model. The ones who have are beginning to widen the performance gap — faster output, more consistent quality, better use of the team's actual expertise.

The question for every Public Affairs leader right now is not "should we be using AI?" It is "what is stopping us from using it properly?"

Usually, the answer is the same thing that holds back most internal Public Affairs improvements: the absence of structure, process, and shared standards. The key right now is governance and methodology — and the willingness to treat AI adoption as an organisational capability question, not just a technology one.

Advocacy Strategy works with in-house public affairs teams on AI integration, PA management consulting, and capability development. If this is a live challenge for your team, get in touch.