On Day 1 in Copenhagen, teams built working AI agents in a single session. Helping people use them wisely? That’s the next build, and it’s a bigger one.

Over at our booth, nobody needed a pitch on what AI can do. What partners wanted to talk about were the humans on the other end of the keyboard: are they ready, and how do you get them there?

It's a question we're a little obsessed with.

At usecure, we're pioneering Human Risk Intelligence (HRI) for MSPs. So while the halls buzzed with agents, automation and managed intelligence, we kept coming back to one thing: where do the people fit?

The big questions on the expo floor

Three themes kept popping up, and they show just how far ahead MSPs and MIPs are already thinking.

First up: readiness. Clients' staff are already using AI tools, and partners see a real opportunity in helping them use those tools with confidence and care, rather than crossed fingers.

Second: sprawl. New tools and agents are turning up faster than anyone can keep a spreadsheet of them. Partners want to stay firmly in control: what's in use, who's using it and what it can reach.

Third: timing. Partners don't want to wait for the first AI-shaped incident to start the conversation. They want clients ready today for whatever tomorrow brings.

Agents take an afternoon. Culture takes years.

Two sessions on Day 1 showed exactly why those themes matter.

At the Microsoft Pre-Day, Microsoft-certified trainers ran an Agent Buildathon. Think hackathon, with a scoreboard. Teams used Copilot Studio and Power Automate to build customer-ready copilots and workflow agents, in a session open to business and technical roles alike. The takeaway: building an agent is no longer a specialist's job. Your account manager could probably knock one up before lunch.

Elsewhere, 'Women in Tech: Building Workplaces Where Everyone Thrives' tackled something no lab can speed up. Rather than another panel, it was a hands-on working session on psychological safety, inclusion and what makes talented women stay, grow and lead. Pax8's takeaway: however fast technology moves, people are still the foundation of every successful business.

Put those two side by side and you've got the challenge facing every MSP right now.

AI tools can land in a week. The habits that keep them safe take a lot longer to build.

Psychological safety is a security control

The Women in Tech session framed psychological safety as a culture issue. We'd say it's a security issue too.

Here's how it plays out. Someone clicks a dodgy link or pastes client data into the wrong AI tool, and the clock starts ticking. If they're worried about getting blamed, they close the tab and hope for the best. If they trust their team, they speak up, and you get precious time to respond.

That matters even more as AI spreads, because mistakes happen faster and are harder to spot. A culture where people put their hand up early is one of the few controls that scale as fast as the tools do.

Three moves to make before the next agent ships

None of these need new technology. They need a decision, and ideally a trial run on your own team first.

1. Set the AI ground rules before the sprawl sets them for you. Agree which tools are approved, what client data never goes near them and who signs off a new agent. Write it down, share it with every client and revisit it every quarter.
2. Measure reporting, not just clicking. Track how quickly people flag a suspicious email or an AI slip-up, like client data pasted into the wrong tool. Time to report tells you whether your culture is working, and it's a number clients instantly get.
3. Rehearse an AI incident with client leadership. Run a 30-minute tabletop exercise: a deepfaked voice note from the finance director asks for an urgent payment. Who spots it? Who speaks up? How fast? Nothing gets clients taking tomorrow's risks seriously quite like watching one nearly happen today.

The question we took into Day 2

One more question came up hot on the heels of the other three: when every user carries some risk, which ones should you worry about first?

That's exactly what Human Risk Intelligence answers, and it's why we built uHealth, the newest part of the usecure platform, to quantify, prioritise and prove human risk.

Day 1 was a good reminder that the groundwork goes in first. Agents will keep getting easier to build, but the people using them are still where security is won or lost.