I can’t believe we’re already wrapping up 2025. What a year it’s been for AI.

So much change, so many updates, and honestly, some developments that completely surprised me.

Things like the quality of AI video now (have you seen Veo 3.1) and Google Gemini’s latest model? Outrageous.

The incredible improvements in text-to-image, agents and of course, all the speculation around AGI.

I thought the best way to close out the year would be to look ahead.

Do I have a crystal ball? Absolutely not. But I’ve done my homework-combed through the research, the expert commentary, and the predictions floating around-and brought it all together through my strategic communication lens.

No matter what kind of business you’re in, communication matters.

And understanding where AI is heading can help us all be more prepared, more strategic, and less overwhelmed by the constant changes.

Let’s dive into my predictions for 2026.

Prediction 1: Budgets Will Shift from Hype to Value

Right now, through my work with strategic communication agencies and teams, I’m seeing a lot of people who are very tool-focused.

They’re jumping at the latest shiny AI tool rather than being strategic about it.

I’ll be honest, myself included. Many of us have been wasting money on tools because we’re not thinking strategically enough.

I think that mindset is going to shift. People won’t spend money on tools unless they can see a clear return on investment and understand how it’s actually going to work for their specific needs.

There’s also been a huge gap in AI training within organisations. Senior management knows they need to integrate AI into their practices, but they’re not training their staff to do so. It’s really interesting from a researcher’s point of view, and I think this training gap is finally going to start closing in 2026.

AI Agents and Agent Lakes

Now, I would say 2025 was the year of the agent, but I don’t think we’ve seen anything yet. AI agents-tools that can go off and do tasks autonomously-are going to become even more powerful.

I’ve used agents to do competitor analysis, where I give them a list of companies, and they go off, analyse websites, and come back with insights in 10 minutes.

Because agents need a lot of data to do their job well, we’re going to see the emergence of “Agent Lakes”-secure repositories of internal data that agents can draw on without sending sensitive information externally.

Prediction 2: Proof Beats Promises

There have been two camps with AI adoption: those who’ve been cautious (which is fair enough), and those who’ve thrown themselves in without evaluating whether it’s actually working.

That’s going to change. AI isn’t a novelty anymore. It’s becoming everyday and necessary. Communication budgets will increasingly link AI use directly to measurable impact. No more AI for the sake of AI.

Inference Workloads Will Increase

When we talk about inference workloads, we mean setting up AI to do specific, repeatable tasks-like writing media releases where you just give it the details and it produces the output. This is different from just using AI for brainstorming here and there.

The more people learn to train AI on their own material, the more this will take off.

Train it on your content, your voice, your brand-and then it can generate work that actually sounds like you. Of course, you still need to fact-check and edit, but it’s so much more efficient.

Evaluation Becomes Essential

I wrote a whole chapter on this in my AI book ‘Artificial Intelligence for Strategic Communication’ because it seems to be missing across the board. Benchmarking is crucial. If you’re using AI to write blog posts, are you measuring how those posts perform compared to when you wrote them all yourself? It’s the same with social media content, is your audience engaging in the same way?

And don’t forget about your customers. I read about a business that replaced their phone number with just a chatbot, but their target audience was seniors who hated chatbots. It’s not about jumping at AI-it’s about thinking strategically about who you’re trying to connect with.

Implications for you:

• Build dashboards to measure AI impact on media engagement and social performance.
• Make AI training ongoing. I recommend monthly sessions, not just once a year
• Master agent operations and learn what autonomous AI can do for your specific tasks.
• Keep developing your human skills-empathy, ethical judgement, and genuine connection will always be your strength.

Prediction 3: Shopping Will Change Completely

This one’s a bit of a fantasy of mine. Imagine telling your AI agent: “Here’s my family. Here’s my budget. Go find Christmas presents for everyone.” It scours the internet, considers delivery times and specifications, and brings back options-or eventually, just makes the purchases for you.

ChatGPT is already about to integrate PayPal, so purchasing within AI tools is coming. I’m still a bit hesitant about an agent purchasing on my behalf (can you imagine linking a credit card with a big limit? Recipe for disaster.), but I think it’ll become commonplace.

Already, nearly one quarter of AI users are relying on AI assistance to find products and services. And here’s the kicker: websites aren’t just going to be visited by humans anymore-they’ll be visited by AI agents on behalf of their humans.

Generative Engine Optimisation (GEO)

This is like SEO, but for generative AI tools. It’s about making sure you can be found when people search in ChatGPT, Gemini, Claude, and other tools.

From the research I’ve read, a few things help:

• Recency of content: AI tools tend to crawl and grip onto recent content. One study showed that simply updating the dates on old content increased discoverability.
• Consistency: Having the same information about yourself across multiple places on the internet-review sites, your website, directories.
• Localised content: Include your local area, even if you service people globally. Remember that example where a woman searched for a social media expert in Coolum and found me? That’s the power of local specificity.

Truth Architecture

This is a term being used to describe processes that prevent misinformation from going out of your organisation. Fact-checking, editing heavily, being that human in the loop-all essential to prevent inaccurate AI-generated content from getting published.

---

Want to hear me walk through all of these predictions in more detail? Watch the full webinar recording below where I dive deeper into each of these trends and share some stories from my recent travels and research.

Now that you’ve had a chance to watch the full discussion, let me cover a few more predictions that I think are particularly exciting for communicators.

---

Prediction 4: Vibe Coding Opens New Doors

If you don’t know what vibe coding is, it’s essentially the ability for non-developers to create apps using AI. You describe what you want, and the AI builds it for you.

I’ve played with this in Canva and other tools. I created an online quiz for my students just by giving the AI my questions and describing what I wanted it to look like. The code gets written right before your eyes.

Previously, we’d pay thousands of dollars to have an event app developed. Now? You could create it yourself.

A word of caution though: apps need data, so be careful about where that data goes and whether it’s sensitive. But honestly, the world is your oyster here. Don’t try to think of something that’ll make you a million dollars. Think of something that’ll save you 20 minutes a day.

Prediction 5: Skills and Ethics Become Differentiators

Prompt writing is still an important skill. Yes, it’s getting easier to tell AI what you want, but the more context and detail you provide, the better your results. It’s like giving instructions to a new team member. Vague input equals vague output.

Teams are going to stop looking at AI for isolated use cases and start redesigning entire processes around it. We’re moving toward true human-AI teams, where some companies have each employee managing 10 agents that handle specific parts of their job.

Prediction 6: Risk and Regulation Increase

The Australian government just announced an AI Safety Institute, which is really exciting. It’s going to collaborate with business owners, scientists, and various stakeholders to help keep everyone safe from misinformation, deepfakes, and harmful AI-generated content.

Looking at us individually, we need to be careful too. I’m still hearing stories of people putting sensitive information into the free version of ChatGPT, even when there’s an AI policy in place. Critical thinking skills around AI need to be tested and developed.

Legal claims are also going to rise-and it’ll be interesting to see how the Australian government’s recent stance on AI tools using copyrighted material plays out.

Implications for communicators:

• Keep building skills and teach others to be responsible with AI.
• Define clear roles in your human-AI teams, with humans always providing ethical judgement.
• Work closely with legal and IT departments on compliance.
• Do safe pilot projects before rolling out new tools organisation-wide.
• Make training ongoing-what you learn today scaffolds what you’ll learn tomorrow.

What I’ve Got Coming Up

I’m so excited about a few things happening soon.

On January 19, I’m launching my six-week course, AI Integration for Communication Leaders. For reading this far you will receive a 40% discount by using the code: AI4COMMS Move quickly on this one. There are only 15 spots available.

We’ll work through everything-policy development, task mapping, process creation, training plans-so you walk away with a detailed, research-led strategy. If you register for the Lightning Lesson, you’ll get a significant discount on the course.

And don’t forget about my free Facebook group the AI for Strategic Communication Community where we talk about AI in the context of strategic communication. I go live every Friday night to chat about the AI news of the week, and community members get discounts on workshops and events.

Looking Ahead to 2026

In 2026 I predict that we’re going to become much more sophisticated in our use of AI. We’ll be more business-focused, linking outcomes to AI use rather than just playing with tools.

We’ll understand and use AI agents more effectively. We’ll fill that training gap and build our skills regularly. And hopefully, we’ll be much more focused on protecting trust and using AI responsibly.

Now I’d love to hear from you: What’s your biggest AI challenge heading into 2026? And what do you think will be the biggest change we’ll see?

If you want to get ahead of the curve, I’d love to help. Book a chat with me.

Let’s talk about how AI can work for your specific situation

Thank you so much for being part of this journey with me in 2025.

Here’s to an exciting, strategic, and responsible 2026.