Paul Zimmerman
INV Group Chief Communications Officer
18th February, 2026
More than a decade ago, I gave a TEDx talk in Woking on what I called the micronisation of communication in the workplace.
My argument was fairly simple: communication was becoming smaller, faster, more distributed, more transient and more ambient. Email was beginning to lose ground to chat. Hierarchies were being flattened by digital collaboration. Language itself was becoming more compressed, more visual and more immediate. The enterprise was starting to behave less like a filing cabinet and more like a living network.
Looking back, I was not far off.
If anything, I understated the scale of the shift.
Today’s workplace is not just built around micronised communication. It is built around continuous communication. Messages, comments, reactions, alerts, summaries, edits, mentions, notifications and workflow triggers now form the nervous system of modern work. Microsoft’s 2025 Work Trend Index found that employees are interrupted every two minutes during core work hours, or 275 times a day, by meetings, emails or chats. That is not a side effect. It is the environment.
That matters because communication is no longer just how work is discussed. In many organisations, communication is now how work is coordinated, recorded, delegated and executed.
That is a much bigger change.
We did not just move from email to chat
One of the clearest trends since my original talk is that the workplace has shifted away from the inbox as the sole organising principle of knowledge work.
Email still matters, of course. It remains essential for formal communications, external engagement and many regulated processes. But internally, the centre of gravity has moved. Slack now describes itself as an AI work platform and the operating system for work, bringing together people, processes, data, agents and AI in one conversational interface. Zoom is doing something similar, turning meetings, summaries, chat and workflows into a more unified layer of work.
That is not just a product story. It tells us something important about what happened to workplace communication.
We did not simply move from one tool to another. We moved from messages as correspondence to messages as infrastructure.
A chat thread is no longer just a conversation.
It may also be a decision log.
A task queue.
A knowledge base.
A project update.
A compliance record.
A trigger for automation.
That is the real evolution.
The old promise was speed. The new challenge is overload
Back in the mid-2010s, it was easy to be excited about all this. Faster communication looked like better communication. More fragments looked like more engagement. More platforms looked like more democratic participation.
Some of that was true.
Digital communications did flatten organisations in meaningful ways. They made it easier for ideas to travel laterally, not just vertically. They made it easier for employees to react, challenge, suggest and contribute in public. They also helped organisations operate across geography, time zones and hybrid work patterns. Gallup’s research shows that hybrid work has proven durable, and that most remote-capable employees still want flexibility rather than a full-time return to the office.
But the more mature view in 2026 is this:
Micronised communication creates the possibility of engagement, but not the guarantee of it.
In some organisations, it has indeed increased connection, responsiveness and shared awareness.
In others, it has created clutter, fragmentation and performative busyness.
Asana’s 2025 research argues that knowledge workers spend 60% of their time on “work about work” rather than skilled work itself: chasing updates, switching tools, attending unnecessary meetings and hunting for clarity. That is a useful corrective. The problem is no longer simply whether people can communicate. It is whether they can communicate without drowning each other in coordination overhead.
This is where my old thesis needs an update.
The question is no longer:
How do we make workplace communication smaller and faster?
It is now:
How do we make it smaller and faster without making work more chaotic, interrupt-driven and cognitively expensive?
Hybrid work made asynchronous communication more important, not less
Another major change since my TEDx talk is the rise and stabilisation of hybrid work.
That matters because hybrid work did not just change where we work. It changed the shape of communication.
In an office-heavy world, a great deal of context lived in rooms, corridors and informal conversations. In a hybrid world, that context has to be made more visible and more portable. It has to survive across time and place. That means more written updates, more comments, more recordings, more summaries and more shared artefacts that people can return to later. Gallup’s research suggests the hybrid era has largely settled into a durable pattern, with flexibility and team coordination now central management challenges.
So the future was never just synchronous chat.
It was always going to be a blend: live meetings, persistent channels, shared docs, recorded video, comments, summaries, search and structured workflow.
In other words, the real shift was not just toward faster communication. It was toward better orchestration between synchronous, asynchronous and now machine-assisted communication.
This is where AI changes the picture again
If the first workplace communication revolution was digitisation, and the second was micronisation, the third is now agentic mediation.
That sounds technical, but the idea is straightforward.
AI is no longer only helping us write better sentences or summarise longer meetings. It is increasingly sitting in the flow of work itself: drafting messages, translating intent, finding context, summarising threads, extracting actions, creating tasks, routing requests and in some cases acting on our behalf.
Microsoft’s 2025 Work Trend Index framed this as the rise of agents that can reason, plan and act as digital labour. Microsoft’s broader 2025 messaging around the “Frontier Firm” goes even further, arguing that organisations are beginning to assemble hybrid human-agent teams. Zoom has expanded AI Companion with agentic capabilities to identify and execute tasks, manage meetings and reduce manual work. Slack is explicitly positioning agents and AI as part of the work operating system.
This is the real inflection point.
For years, workplace communication was about humans exchanging information more efficiently.
Now, communication is increasingly becoming machine-readable instruction.
A manager’s message may become an automated workflow.
A meeting discussion may become an AI-generated action list.
A support query may become a routed case.
A project update may trigger follow-up requests and reminders.
A policy question may be answered by an internal agent trained on organisational knowledge.
The communication itself is no longer the end of the process.
It is the beginning of machine action.
That changes the meaning of communication in the workplace.
AI agents may make work more conversational — and more operational
One way to think about AI agents is that they collapse the distance between saying and doing.
Historically, workplace communication often involved several steps.
A person asked a question.
Someone else interpreted it.
A task was created.
The task was passed on.
A document was updated.
A status report was circulated later.
Agentic systems compress those stages.
Increasingly, you will be able to say:
“Summarise the key risks from this thread, draft a response for legal review, open actions for the team, and brief me before tomorrow’s meeting.”
And the system will do at least some of that.
That has profound implications for workplace communication.
It means that communicative skill will increasingly include the ability to:
- express intent clearly
- define boundaries
- give usable context
- review machine outputs critically
- know when not to automate
The best communicators in the AI workplace will not just be persuasive or articulate. They will be able to communicate in ways that are legible both to humans and to systems.
That is a new form of organisational literacy.
But agentic communication also raises the stakes on trust
This is the part leaders should take seriously.
As soon as AI agents begin acting within workplace systems, communication stops being only a cultural matter. It becomes a governance matter.
If an AI agent can summarise a meeting badly, that is annoying.
If it can assign the wrong action, route the wrong information, misstate a policy, leak sensitive context or act without sufficient oversight, that is a different class of problem.
NIST’s AI Risk Management Framework remains relevant precisely because AI risks are not just model risks. They are organisational risks. This week, NIST announced an AI Agent Standards Initiative focused on secure interoperability and confidence in agents capable of autonomous actions. That is a strong signal that the standards and assurance world sees agentic systems as qualitatively important, not merely another software feature. Meanwhile, Gartner warned earlier this month that by 2028 misconfigured AI could shut down national critical infrastructure in a G20 country.
That may sound extreme, but the underlying lesson applies just as much inside ordinary enterprises.
The more communication becomes executable, the more it requires:
clear permissions,
good audit trails,
strong identity and access controls,
human override,
role clarity,
policy boundaries, and
proper evaluation of how agents behave in real-world conditions.
In short, the future of communication is not only about speed and convenience.
It is also about control, assurance and accountability.
The real risk is not that AI agents replace communication
It is that they make bad communication scale faster.
If an organisation is already unclear, overly political, badly coordinated or weak on governance, AI agents will not fix that on their own. They may amplify it.
Poorly framed prompts will produce poor outputs.
Ambiguous policies will create inconsistent behaviour.
Messy systems will generate messy automations.
Weak management will become faster weak management.
This is why the next era of workplace communication will belong to organisations that combine two things well:
clarity of human intent and governed machine execution.
That is the winning combination.
Not “more AI”.
Not “more messages”.
Not “more productivity theatre”.
But clearer thinking, better operating models and trusted systems that help people move from discussion to action without losing judgment along the way.
So where does this leave the workplace?
I still believe the core idea from my TEDx talk stands.
Communication has indeed become smaller, faster, more distributed and more ambient.
But in 2026, that is no longer enough as a description.
Today’s workplace communication is also:
- persistent
- searchable
- multimodal
- asynchronous
- interruption-heavy
- AI-assisted
- and increasingly agent-mediated
That means the modern enterprise is not merely a network of people speaking to one another.
It is becoming a network in which humans, systems and agents continuously exchange signals, interpret context and initiate work.
That is a very different kind of organisation.
The challenge for leaders now is not simply to install better communications tools.
It is to build a workplace in which:
communication is clear,
coordination is intentional,
AI is governed, and
agents operate in service of human judgment, not in place of it.
Because the future of the workplace will not be defined by who communicates the most.
It will be defined by who can turn communication into trusted action most effectively.