Focus: From Investment to Impact | January 2026 Newsletter

1. The Month in View: 3 Strategic Trends

Curated insights to help you read the room and lead the shift.

Trend 1: AI Investment Is High – Returns Are Not

The PwC Global CEO Survey made for sobering reading as we kicked off the year. Confidence in AI is still high – most CEOs are increasing their AI investment – but the number of organisations that can point to measurable returns remains stubbornly low. That gap isn't about technology. It's about the fact that deploying tools without changing how people work, how decisions get made, or how capability is built tends to generate activity rather than impact. Gartner's research into what AI in the workforce actually means reinforces the same point: the organisations seeing real returns are the ones that have treated this as a human change programme, not just a technology rollout.

Trend 2: The People Agenda Moves to the Centre

One of the clearest signals heading into 2026 is that the organisations who get most from AI will be the ones who invest seriously in their people. Forbes' HR trends piece and BCG's research on the AI skills gap both point in the same direction: the capability gap is real, it's widening, and it won't close on its own. What's interesting is that the gap isn't mainly about advanced technical skills – it's about the foundational stuff: knowing how to work with AI effectively, how to evaluate its outputs, and how to build that into everyday practice. That's where microlearning, done well, has a genuine role to play.

Trend 3: Getting Ready for the Agentic Future

The World Economic Forum's piece on agentic AI is worth sitting with. The argument isn't that agents are a distant prospect – they're already here in limited forms – but that most organisations haven't yet thought clearly about what they want AI to do autonomously, what decisions they're comfortable delegating, and what governance that requires. Cambridge Judge's look-ahead piece makes a similar observation: the organisations that will fare best in 2026 aren't necessarily the ones moving fastest, but the ones with the clearest sense of purpose and the strongest foundations underneath their ambition. As we keep saying: AI works best when humans have done the hard thinking first.

Claude Code

Launch: February 2025

For most of AI's short history, you still needed a developer to build anything. You could describe what you wanted, but someone had to translate that into software. Claude Code changed that. For the first time, you could describe an outcome – a tool, a report, an automated process – and the AI would build it, test it, fix its own mistakes, and deliver working software without hand-holding. Within months every major AI lab had shipped a competing version. By November 2025 it had reached $1 billion in annualised revenue, one of the fastest product trajectories on record.

Value: Internal tools that would previously have required an IT project and a six-week queue are now being built in an afternoon. Organisations that grasped this in 2025 are moving faster in 2026. The question for your team is whether you're one of them.

GPT-5.2 Solves Unsolved Mathematics

December 2025

In December 2025, GPT-5.2 generated original proofs for mathematical problems that had been unsolved for decades – verified as correct by one of the world's leading mathematicians. The problems were at the accessible end of the unsolved set, and the mathematician was careful to say so. But the point stands: an AI produced original expert-level reasoning that no human had managed. That is a different kind of capability from writing emails faster.

Value: If AI can do original expert work in mathematics, the question of whether it will reach that threshold in law, finance, strategy, or your sector is no longer hypothetical. It is a matter of when. The organisations building AI literacy now will be better placed to use that capability – and govern it – when it arrives in their domain.

Deep Research

OpenAI February 2025 · Google March 2025 · Claude June 2025

You ask a question. Twenty minutes later you get a structured, cited report that would previously have taken an analyst the better part of a day. Competitive landscape, regulatory context, market sizing, policy analysis – the kind of background work that used to mean waiting until tomorrow now completes while you're in your next meeting. Deep Research is now standard across every major AI platform, and we use it regularly at Caversham House. The output needs a careful eye, but the time saving is real and immediate.

Value: This is the most accessible entry point into agentic AI for most business users. Try it with a real question from your work this week. That is the fastest way to understand what has changed.

Computer Use

Anthropic August 2025 · open beta December 2025

Most AI tools only work with software that has been specifically connected to them. Computer Use takes a different approach: the AI watches your screen and operates your browser the way a person would – clicking, typing, filling in forms, moving between tabs. By December 2025 Anthropic had made this available to all paid subscribers as a simple Chrome extension. Think of the repetitive web-based tasks in your week: updating systems, pulling information from multiple sites, completing routine processes across tools that don't talk to each other. That is the territory this opens up.

Value: The capability is real and improving quickly. So are the risks – an AI acting on your behalf in your browser needs clear boundaries and oversight. The organisations that will use this well are the ones building governance frameworks now, not after the fact.

Ready to take the next step?

Already a subscriber? Look out for our new microlearning modules next month! They will appear live in your training portal.
Thinking about subscribing? Take a look at our 2026 Training Subscriptions to see what's included and how it works for your team.
Want to talk strategy? Book a briefing or contact us and we can work through your 2026 AI priorities together.