Bombos teamBy Piet Baudoin · 22 September 2026
AI is showing up more often in ordinary work, replacing people can turn out costly, and Meta’s Muse asks for your approval before sensitive actions. Today: what these three stories mean for your employees and the work that comes back every week.
How many people are using AI at work?
Commercial roles now use AI more intensively than IT does, Computable, a Dutch IT trade publication, reported today. The article discusses the AI in Nederland 2026 survey by Maurice de Hond and Vincent Everts among roughly 6,000 people in the Netherlands.
“Commercial roles are now in the lead with 81 percent intensive use, followed by education roles (77 percent)” (translated from Dutch). Among people in the Netherlands as a whole, use keeps growing: the share who never use AI fell from 55 percent in January 2024 to 28 percent in September 2026.
Those numbers do not tell you how your own business works. They do make it likely that your employees are already using AI, each on their own laptop. That is not yet the same as a fixed way of working for the business.
Ask your people what they use AI for. Then look at which recurring task you could organize better together, for example keeping email with the right client file.
Source: Computable, 22 September 2026
Can I use AI to do the same work with fewer people?
That can turn out costly later, Gartner warns, according to Computable. The prediction: “by 2029, almost 30 percent of employees who were laid off because of AI will have to be rehired” (translated from Dutch).
Note that this is a forecast, not a measured outcome. Gartner is a research firm that itself sells advice on this subject. Its advice is to strengthen your existing employees rather than lay off large numbers of them.
For a business, the more relevant question is closer to home: which tasks can you actually hand to AI? Your employee does more than one thing. If the groundwork for one task disappears, their other responsibilities are still there. The gain then sits in more work of higher quality with the same people.
Source: Computable, 21 September 2026
Does an AI assistant ask permission before it does something?
Meta brought its assistant Muse to the Mac, TechCrunch reports. It can “interact with your files, messages, calendar, notes, and mail, all within their native applications.” In plain terms: it works with your files, messages, calendar, notes and mail, inside the programs you already use.
You choose what it can reach: “you can control what Muse has access to on your Mac on an opt-in basis.” You grant access deliberately.
For sensitive actions: “the app will always ask for your approval before it performs sensitive actions.” So it asks for your approval first. That does not mean every action asks for permission. It does show how human approval is becoming a fixed part of assistants that carry out work. For email, that raises the practical question: should an AI assistant send email on its own?
Source: TechCrunch, 18 September 2026
What does this mean for your business?
Start with one recurring task that currently gives your people a lot of groundwork. Agree on what AI gets ready and who reviews it.
Bombos takes you through that, so you can do it yourself afterward. If we start with a shared inbox, our promise is this: every morning, the replies are ready for review and no email is left sitting. Bombos gets it ready, a person approves it. Sending only happens after that approval.
Sources
Every source was opened on 22 September 2026 and every quote appears in it word for word.
Ai-revolutie verovert de werkvloer: commercie en onderwijs halen ict in (22 September 2026) · Ai-ontslagen kunnen bedrijven duur komen te staan (21 September 2026) · Meta’s Muse hits Mac, letting the AI take actions on your computer (18 September 2026)