AI in 2026: what we’re watching out for
As a CEO building software platforms that organisations rely on day in, day out, I’m approaching 2026 with a mix of excitement and responsibility. AI is no longer something we talk about in the abstract — it’s becoming part of core operations, and that changes the questions we need to ask.
From my perspective, these are the things that will really matter in AI over the next year or so.
1. AI stops being a feature and starts being part of the workflow
As a business leader, I’m far less interested in AI as a standalone feature than I am in how it shows up inside real workflows.
Right now, chat is still the dominant interface. People ask a question, get an answer, and move on. In 2026, the shift I’m focused on is AI that operates inside the flow of work — agents that run on schedules, monitor activity, and produce work that teams can validate or approve.
That’s where AI stops being impressive and starts being useful.
2. Agentic AI needs management, not blind trust
We’re investing heavily in agentic approaches, but I’m very clear on one thing: full autonomy isn’t the current goal for most businesses.
From a CEO’s point of view, the idea of setting objectives and “leaving it to the agents” is risky. Efficiency without context can produce outcomes that are technically correct but operationally suboptimal.
What I’m excited about is what I call lane assist - agents that support teams, keep things on track, and surface issues early, without removing human oversight.
3. The real opportunity sits in the middle ground
At one end of the spectrum, we have simple Q&A bots. At the other, fully autonomous systems making decisions.
The space I care most about as we head into 2026 is the middle: AI as a cognitive partner. Systems that understand our processes, our SOPs and our context, and that proactively surface information rather than waiting to be prompted.
For a CEO, that’s incredibly powerful - it means fewer blind spots and better-timed decisions.
4. AI becomes a work amplifier across the organisation
I increasingly think about AI as a work amplifier rather than a replacement technology.
In large, complex organisations, no single person can see everything that’s happening. In 2026, I expect AI to play a much bigger role in bringing critical information to the surface — especially in operational, case-based and high-volume environments.
From a leadership perspective, that’s about scale, consistency and resilience, not headcount reduction.
5. We finally get honest about model training
One thing I hear a lot is, “we trained a model”. Most of the time, that’s not actually true.
What I’m watching for in 2026 is better tooling that makes genuine model training achievable for more organisations. Smaller, open-weight models trained for specific purposes can often outperform large, expensive models — and do so more efficiently.
As someone responsible for both performance and cost, that matters.
6. Efficiency becomes a leadership issue, not just a technical one
As context windows grow and AI becomes cheaper, it’s tempting to just use more of it.
I don’t think that’s sustainable.
In 2026, I expect more focus on how efficiently AI is being used. Choosing the right model, minimising unnecessary compute, and thinking about environmental impact are no longer purely technical concerns — they shape how responsibly organisations deploy AI.
7. Security in an agentic world becomes non-negotiable
Once agents are making decisions and touching real systems, security stops being theoretical.
From a CEO perspective, one of the biggest challenges heading into 2026 is understanding how all these pieces are bolted together — where data flows, where decisions are made, and how risk is managed when agents are part of the operational fabric.
Agentic AI raises the stakes, and security has to evolve alongside it.
8. The real value is organisational wisdom, not AI judgment
The thing that excites me most looking ahead isn’t AI making judgments on our behalf.
It’s AI helping organisations move from knowledge to wisdom — making sense of vast amounts of information that no human team could realistically process alone.
AI may make judgments at some point, but the most valuable part for me is improving access to knowledge and building organisational-level wisdom that supports better decision-making.
Looking ahead to 2026
From where I sit as a CEO, the next phase of AI is about maturity.
The organisations that do well won’t be the ones chasing autonomy for its own sake. They’ll be the ones embedding AI into workflows thoughtfully, treating agents like colleagues rather than magic boxes, and focusing on amplification rather than replacement.
That’s what I’m watching as we head into 2026 — and that’s where I believe the real value will be created.
Get started with confidence - download our AI guide