Calls to slow down the pace of artificial intelligence development, driven by fears about its potential to threaten humanity, are unlikely to resonate with most office workers. After all, the majority of them still use the technology for pretty basic tasks. According to a Deloitte survey of 25,000 employees in the UK, published this month, only about a quarter of British adults use AI tools every day at work. And in most cases, the usage remains fairly simple: searching for information, writing emails, fact-checking, and creating summaries. 🤖
But there is a smaller group that decided to play on a whole different level. These people have restructured their entire professional lives around the intensive use of new tools. The Microsoft Work Trend Index 2026 calls this group AI power users, and they represent just 16% of artificial intelligence users in the workplace. The report identifies three characteristics that define this profile: they are advanced users of AI agents, they have redesigned workflows to maximize efficiency, and they share what they know with the rest of the organization.
But what does it actually mean to be an AI power user at the office? Executives from companies like Microsoft, Notion, Lewis Silkin, Oliver, Syspro, and BCG shared their experiences, and the stories show that this goes way beyond asking for an email summary. 💡
Feed Your Agents With Institutional Knowledge
For some leaders, an AI agent works almost like a chief of staff. Pamela Maynard, director of AI transformation for client and partner solutions at Microsoft, describes her agent as someone who holds institutional context: her priorities, her schedule, what she truly values, and even her personal boundaries. She says the agent uses all of this to prepare her, shaping her week, highlighting what needs her attention, and making sure she walks into every meeting ready to go.
And there is a curious detail. According to Pamela, the agent also challenges her decisions. On one occasion, it actually asked why she was working on a Sunday afternoon when she should be spending time with friends or family. This kind of interaction shows how agents have moved beyond being passive tools and now act proactively within the routines of the people who use them.
Angela Tangas, who leads the British marketing agency Oliver, part of the digital advertising and AI group Brandtech, created an agent called Ask Ange. It is not an avatar yet, but it works as a first line of support for her team when she is unavailable. According to Angela, it all started with the most common questions she gets when talking to clients, the way she typically answers them, and where she goes for sources. It is a work in progress: if the question is too complex, the agent simply wraps up and directs the person to the real CEO.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
Redefine What You Are Capable of Achieving
At the law firm Lewis Silkin, senior associate Jordan Quartey uses the specialized legal tool Harvey to pull together client-specific files containing the firm’s own past opinions and decisions, along with current email exchanges. With that, he no longer needs to turn to more experienced colleagues right at the beginning of a case. He can make significant progress on his own, learning from what has been done before and using that context to develop his own judgment. In his words, it is more efficient and also a better way to learn. ⚡
This is exactly the kind of gain that defines a power user: the technology does not replace the professional, but frees up their time and expands what they can accomplish independently.
Ivan Zhao, cofounder and CEO of Notion, the collaboration and productivity platform, uses a really great musical metaphor to explain the shift in work patterns. He compares the difference between a marching band and a jazz ensemble. The marching band, he says, is like the old way of working: everyone in synchronized formation, following the same sheet music as always, burning a ton of energy just to stay aligned.
The jazz ensemble, on the other hand, has talented people improvising around a shared sense of where the music is heading. In his work environment, that means agents handle the boring and repetitive stuff, while the team spends its time on improvisation and fine-tuning. According to Zhao, this leads to more long conversations between colleagues, discussing what they are building, why they are building it that way, and what they should do next.
Prioritize Your Human Colleagues
For Maynard, at Microsoft, AI should give us more room for the human element, not less. The time saved through efficiency gains needs to be spent on relationships, mentorship, and those messy human conversations that no agent can have on our behalf. She makes an important point: the real risk is letting the recovered time get silently filled with more tasks.
While some leaders just want to cut meetings, Maynard prefers to make them more worthwhile. An agent handles the pre-briefing so that teams can use the time for the real conversation. The summary and next steps practically write themselves, so the time in the room is dedicated to people, not process.
Zhao says the Notion team eliminated the boring meetings, because an agent can summarize which products shipped that week or which sales were closed. This opens up space for more spontaneous meetings, the kind where you notice something during lunch or walk by a colleague’s desk and start chatting about what is on their screen. According to him, that is exactly where the best ideas tend to come from. 💬
Stop Thinking of AI as a Time Saver
Angela Tangas has an interesting take on this. She says she does not measure the benefits of her AI agent team by the time it saves, but by how the technology makes her a more effective leader.
She starts her day with an AI-generated news briefing, carefully tuned to surface information about the industries relevant to her business. Angela gives a concrete example: if a client issues a profit warning, she probably would not have caught that without her configured prompt. For her, this allows her to be much more proactive on a day-to-day basis.
Then she works with agents inside the company’s own AI platform to brainstorm ideas and prepare for client meetings. Before meeting an executive, for instance, she describes the client’s profile to the agent, explains what she hopes to accomplish in the conversation, and asks it to simulate how that dialogue might play out. It is almost like an intelligent rehearsal before the real game.
Leanne Taylor, CEO of Syspro, a software manufacturer, reinforces this point by saying that how many hours she saves is the wrong question. Her company’s AI tool is trained on corporate routines and discussions, but it also learns how she personally thinks, decides, and prioritizes. The system understands the language she uses when she is reflecting on something, as opposed to the language she uses when she has reached a conclusion. That means the technology helps her make better decisions that reflect her way of leading, not some generic version of leadership. 🚀
Watch Out for So-Called Bot Sitting
Berry Diepeveen, senior partner at BCG and head of the consultancy’s technology and digital practice for the UK, Netherlands, and Belgium, says he has virtual teams working for him day after day. But he makes a critical caveat: proper oversight is vital.
According to Diepeveen, if he creates a massive amount of content with AI support but simply throws it all over the wall and a human suddenly has to deal with five, ten, or fifteen times more volume of work, then the bottleneck has just moved somewhere else. This warning matters, because it shows that advanced AI use cannot turn into a machine that pushes tasks onto colleagues. The gain needs to be real for the entire chain, not just for whoever is operating the agents.
And Above All, Have Fun
To wrap things up on a surprising note, Leanne Taylor from Syspro sums up what might be the most unexpected takeaway of all. According to her, the most surprising thing about AI is how much fun it is. Taylor expected something useful and efficient, but she did not expect to enjoy the way she works more than she did before. She describes that something happens when friction goes down: you start to play, to experiment, to test without fear.
And maybe that is where the big lesson behind all of this really lives. AI power users did not reach this level by following a rigid playbook or taking a specific course. They got there because they treated the technology as a constant field of experimentation, with the freedom to fail, adjust, and share what they discovered. The most relevant digital transformation happening right now does not start in the IT department. It starts with the behavior of people who decide to use technology in a different way.
Technology without behavioral change is just cost. But when professionals change the way they work, experiment, collaborate, and actively learn, AI stops being a simple tool and becomes a real multiplier of capability. Power users already figured this out in practice. The question that remains is: how long will it take for the rest of organizations to reach the same conclusion? ⚡
