Dealing with AI information overload

We’re mid-way through June’s internal comms event season, and I’m exhausted. 

Last Thursday alone included a breakfast briefing, an afternoon conference and an evening roundtable. This was all on a tube strike day. But it’s not the travel that made me tired, or the 15,000 steps around ExceL London. It’s the sheer volume of content on AI: keynote after keynote offering a perspective, or hot take, on how AI is changing the world of internal comms, as well as every intranet platform vendor explaining why their "added AI" is the best. 

I’ve diagnosed myself with AI information overload.

Not everything is a breakthrough

When everything is positioned as a breakthrough, it becomes harder to see what’s useful. The challenge for us is filtering the noise and identifying what will genuinely transform how we work. The potential of AI is exciting, but there’s a growing gap between the headlines and the reality on the ground.

Let’s take agentic AI, which was the focus of one breakfast roundtable. I was curious to hear how people were actually using it day to day. There was one interesting example of a practitioner who had created an AI agent linked to the org chart, helping leaders understand what AI could do and where the team still added value. 

But beyond that, the reality was far less mature. For most, it’s still early experimentation. Teams have been given the tools, typically Microsoft Copilot, and largely left to figure it out for themselves.

Gartner predicts that by 2028, 75% of employees will get their internal comms from AI chatbots. Given that’s only around 18 months away, my question is: what needs to happen between now and then to make that a reality? Because the truth, from the conversations I’m having with in-house practitioners, is very different. Despite a growing assumption that teams don’t need more headcount because “AI can do it”, the reality is much more investment is needed in training and upskilling.

Hope is not a strategy

As I often say, hope is not a strategy. If organisations are serious about adoption, the evidence is there. For every £1 spent on licences, closer to £4 should be invested in the change around it, including skills, support, governance and ways of working. Yet, in a room full of practitioners, not one had been given any budget.

I can’t manifest budget – but I can offer five practical ways to start bridging the gap.

1. Start with the work that’s already been done 

Professional bodies have done a lot of the heavy lifting. The Chartered Institute for Public Relations (CIPR), for example, has published The Responsible Use of AI in PR, translating the Global Alliance’s Venice Pledge into practical guidance. It’s available to both members and non-members.

2. Step back and map your work

Ask three simple questions: what could be done by AI? What needs humans and AI working together? And what should remain human-led? For the work that could be done by AI, consider whether an agent could take on a specific, repeatable task.

3. If you want to experiment with agentic AI, keep it simple 

Start with one agent and choose a task that would make a tangible difference to your team today. That might be summarising insight, supporting content triage or answering common employee questions.

4. If formal budgets aren’t there, take control of what you can 

Build a learning plan, mixing events, webinars, hands-on experimentation and peer learning.

5. Be intentional about what you pay attention to 

Curate a small number of trusted sources so you can distinguish what’s genuinely relevant from what just feels important because everyone’s talking about it. And don’t underestimate peer networks: the most valuable insights often come from honest conversations about what’s really working, and what isn’t.

I’m starting the international leg of event season next week as I head to Toronto for the IABC World Conference, where I expect AI will be high on the agenda – again!