When AI Joins the Conversation, Who Sets the Rules?


Part 3 of our series exploring what happens to trust when AI becomes part of relationships.

Several years ago, I was on a work call when a colleague introduced an AI note-taking assistant into our meeting.

I remember wondering why we needed another recording tool when our existing meeting platform already offered similar capabilities. I also wondered whether everyone understood what the new assistant would capture and how that information might be used.

Afterward, participants began receiving invitations and notifications from the service. It was a little annoying, but what stayed with me was how easily one person’s decision to introduce AI could change the experience for everyone on the call.

I don’t think anyone intended for that to happen. We were all experimenting with new tools and figuring out what worked.

But it raises a question that’s becoming increasingly relevant for organizations: Who decides how AI participates in workplace conversations, and what expectations should guide those decisions?

When recording becomes personal

Recording and transcribing workplace meetings isn’t new. And AI note-taking offers real benefits, from helping people stay engaged in conversations to reducing the administrative work of documenting decisions and next steps.

I use these capabilities myself and appreciate their value. I also believe in letting people know when AI is recording and asking permission before doing so.

What’s changing is the growing number of ways conversations can be captured and what AI can do with that information afterward. Meeting transcripts can become searchable knowledge, inform future AI-generated work, provide context to other systems and agents, or even generate interpretations about the people involved. And increasingly, the tools capturing that information may belong to individuals rather than their organizations.

A few months ago, my nine-year-old daughter told me that a classmate had recorded her and her best friend using what appeared to be ordinary sunglasses. Her friend recognized that they were actually smart glasses with recording capabilities.

I hadn’t expected to have a conversation with my daughter about AI-enabled recording devices on the playground. Her school already had a smart-device policy, but the technology wasn’t necessarily recognizable to the adults around her.

It made me think about how easily similar situations could arise in workplaces as AI-enabled wearables and personal assistants become more common.

And it raises a broader consideration for organizational culture.

Informal conversations, whether around the watercooler or during everyday collaboration, help people develop ideas, build relationships and solve problems together. Organizations should consider how expectations about recording and AI participation might influence those interactions.

The goal isn’t to discourage useful technology. It’s to make sure people understand when and how AI is participating in their conversations.

The decisions organizations need to make

For many organizations, AI capabilities are evolving faster than the policies and shared practices governing their use.

One employee uses their company’s approved meeting platform. Another prefers a separate AI note-taking service. Someone else uses an application connected to a personal account or an AI-enabled wearable.

Each may be trying to accomplish the same thing more efficiently. But these individual choices can create inconsistent expectations about approved tools, disclosure, data access and information sharing.

Recent research reinforces why shared expectations matter. In October 2026, EQUS released What People Really Want From AI: The Personal AI Report, a vendor-sponsored survey of 1,034 U.S. adults who use AI assistants for personal matters. Among respondents, 65 percent reported sharing information about someone else with an AI assistant, while one in four full-time workers reported sharing information about a coworker or client.

While the findings aren’t representative of all Americans, they highlight an important consideration: responsible AI use isn’t only about protecting the information we choose to share. It’s also about recognizing our responsibility for information involving other people.

That’s why organizations need to consider whether their AI policies and guidelines have kept pace.

Are expectations clear for AI recording in internal and external meetings? Do employees know which tools and devices are approved? How should recordings, transcripts and AI-generated notes be handled? And do existing guidelines address personal AI assistants and wearables?

These are questions that may require input from HR, legal, privacy or security professionals, but recognizing and coordinating those decisions is an important part of AI governance.

Technology providers are also introducing new controls. In 2026, for example, Google introduced an optional administrator setting in Google Meet that allows organizations to require participant consent before recording, transcription or AI notetaking begins.

The technology provides the option. The organization still needs to decide how to use it.

An organization that has articulated its position on human agency, transparency and responsible AI use has a stronger foundation for deciding how AI should participate in workplace conversations.

This is where an organizational AI Point of View (POV) becomes valuable, helping leaders establish principles that guide decisions as new capabilities emerge.

Designing for the work and the relationships

There’s another side to this discussion that I think organizations may be overlooking.

Meeting documentation is a workflow that many teams probably never thought they needed to formally design.

For years, people took notes however they preferred. Someone typed into a document, someone else wrote in a notebook and another person relied on the meeting platform’s recording feature.

The approach might vary, but the implications were generally contained within familiar organizational practices.

Now AI is changing that.

A meeting might involve an AI assistant capturing notes, another tool summarizing decisions and an agent turning those decisions into tasks or follow-up communications.

Without shared practices, what began as individual attempts to work more efficiently can become a patchwork of tools, processes and information-handling decisions.

And the underlying workflow may never have been examined.

What information actually needs to be captured? Who is responsible for reviewing AI-generated summaries? How should decisions and action items be documented? Which approved tools best support the work? And what expectations should apply before AI is introduced into a conversation?

These are questions that connect governance directly to Human + AI workflow design.

At Human Driven AI, this is part of how we help organizations think more intentionally about adopting AI. We assess tools and workflows against organizational needs, establish shared practices and help design how people and AI work together.

The goal isn’t necessarily to add more technology. In some cases, the organization’s existing approved tools may already provide the capabilities employees are looking for.

What may be missing is a shared approach to using them.

And just as importantly, those decisions should consider the kind of workplace culture the organization wants to sustain.

Meetings aren’t simply transactions where information is exchanged. They’re also places where people build relationships, develop ideas, challenge assumptions and learn from one another.

When we design how AI participates in those interactions, we’re influencing more than efficiency. We’re influencing the conditions under which people collaborate.

That’s why I believe intentional AI design needs to consider both the work being performed and the relationships surrounding it.

Preserving trust as AI becomes part of the conversation

I think back to that meeting when an AI note-taking assistant unexpectedly joined our conversation.

Today, there are far more ways for AI to listen, record, summarize and participate in the interactions we have with one another. And with personal AI assistants and wearables becoming more accessible, those interactions won’t always happen inside an organization’s approved meeting platform.

Some of these capabilities can make our work easier and our collaboration better. I certainly value being able to participate more fully in conversations when AI helps capture the important details.

But I also believe we need to be thoughtful about what should be recorded, how people are informed and what happens to the information afterward.

Because the decision to introduce AI into a conversation affects more than the person who activates it. The technology may belong to the person using it, but the interaction doesn’t.

Being enthusiastic about AI and believing there should be places, moments and information that remain outside its reach are not opposing positions.

They’re both part of designing AI’s role in our workplaces with intention, good judgment and respect for the people affected.


Human Driven AI helps organizations move from scattered AI experimentation to shared, scalable practice. We build the governance foundations, redesign workflows around human and AI strengths, and deliver custom training programs that teach your teams how to put those systems into practice.

Whether you need an executive Lunch-and-Learn, a hands-on offsite workshop, an enterprise AI transformation program or a strategy to strengthen your brand’s visibility through GEO, we help you turn AI capability into better ways of working.

Ready to move from using AI to working differently with it? Let’s talk.

Read more

Stop Building AI Agents Until You Understand the Work

Everyone wants an AI agent. But automating a workflow you don’t fully understand can simply make a bad process faster. Before organizations deploy agents, they need to determine what should remain human, where AI should assist, what should be automated, and how teams will actually work inside the new system.


Posted

in

, ,

by