Latest / Future of Work Tech with Fexingo: Remote Tools, AI Productivity, and Workplace Software / How AI Is Writing Your Internal Company Communications
Transcript
- Lucas: So there is this growing trend inside companies that you probably haven't heard much about — and it's AI tools writing the internal communications that managers and HR teams send out. Luna: You mean like Slack messages, email blasts, all-hands announcements — that kind of thing? Lucas: Exactly. I'm talking about a mid-size tech company — around two thousand employees — that started using an AI assistant we'll call 'MemoBot' to draft nearly all of their internal company communications. And the results are pretty revealing. Luna: OK, so walk me through how it actually works. Does a manager just type a few bullet points and the AI writes the whole thing? Lucas: Pretty much. The tool ingests the company's past communication patterns — previous all-hands emails, Slack messages from leadership, even the tone used in different departments. Then when someone wants to send an update, they specify the audience, the topic, and the desired sentiment — for example, 'upbeat but factual' or 'serious and direct'. And the AI generates a draft that matches the company's voice. Luna: And does it actually get used? Or is it one of those tools that gets installed and then ignored after a week? Lucas: The data from this particular company showed that after three months, about sixty percent of all internal communications were at least started by the AI. And the average time a manager spent writing an announcement dropped from about twenty minutes to under five. Luna: That's a huge time savings. But I have to wonder — does it make the messages feel generic? Like, if everyone's using the same tool, doesn't everything start to sound the same? Lucas: That was the initial concern. But what they found was that the AI actually personalized the tone based on the department. Engineering got shorter, more direct messages. Marketing got slightly more enthusiastic language. And leadership messages — like from the CEO — kept a more formal structure, but with warmth dialed in based on past successful engagements. Luna: So it's not just a one-size-fits-all template. It's actually tailoring the message to the audience. Lucas: Right. And there's some interesting nuance there. The AI also learned which subject lines got higher open rates, what time of day messages were most read, and even which emojis — if any — resonated with different teams. Luna: Emojis in internal comms — that's a whole other conversation. But I can see how that level of personalization could actually improve engagement. Lucas: It did. The company reported that open rates for ai drafted messages were about twelve percent higher on average than manager-written ones. And employee survey scores around 'communication clarity' went up by eight points. Luna: But let me push back a little. If a CEO's message is being drafted by AI, does that undermine authenticity? I mean, part of leadership communication is the human touch — the imperfect phrasing, the personal anecdote that doesn't fit a template. Lucas: That's a really fair point. And the company actually addressed that by letting leaders customize the AI's draft — adding their own stories, adjusting the tone. The AI was a starting point, not the final word. But there is a risk that if leaders become too reliant, the messages could feel sanitized. Luna: And what about the really tough messages — like layoffs or reorganization? Can AI handle that? Lucas: That's where it gets tricky. The company deliberately kept AI out of sensitive announcements for the first six months. They felt — and I agree — that delivering bad news requires a level of empathy and personal accountability that an AI shouldn't simulate. Luna: So the AI is for the routine stuff: policy updates, event reminders, project milestones. But when it matters most, a human still writes the message. Lucas: Exactly. And I think that's the right boundary. But I also wonder how long that line will hold as the technology improves. There are already companies experimenting with AI that can detect emotional tone in real time and suggest language adjustments. Luna: That feels like a slippery slope. If an AI is telling you how to sound more empathetic, are you actually being empathetic? Or just performing empathy that a machine optimized for? Lucas: That is the billion-dollar question. And it gets to the heart of what we're seeing across workplace AI: these tools are incredibly efficient, but they also force us to ask what we value about human connection in the first place. Luna: You know, this conversation reminds me why I love this show. We dig into these real-world applications that are quietly reshaping how work gets done. And if today's episode gave you something useful to think about — or maybe a new tool to consider for your own team — I know we both appreciate that. Lucas: Yeah, absolutely. And if you found value in this, the way we keep the show ad-free and focused on real substance is through listener support. You can buy me a coffee dot com slash fexingo — that's buy me a coffee dot com slash fexingo. It helps us keep doing deep dives like this without any sponsor noise. Luna: We really do appreciate that. OK, so back to the topic — one other thing I wanted to ask: what about the legal or compliance side? If an AI drafts a policy update, who's responsible if the wording is ambiguous or even incorrect? Lucas: That's a huge issue, and it's one reason some companies are slow to adopt. In this case, every ai generated message went through a human reviewer before sending. But as the tool gets more autonomous, companies will need clear governance policies. Some are already adding disclaimers — like 'this message was drafted with AI assistance' — to manage expectations. Luna: So transparency might be part of the solution. Let people know an AI helped write it, but a human signed off. Lucas: Exactly. And I think that balance — leveraging the efficiency of AI while preserving human judgment — is the model that will win in the long run. It's not about replacing the communicator; it's about giving them better tools. Luna: Right. So the future of internal comms might be a hybrid: AI handles the draft, humans handle the soul. Lucas: I love that framing. And if that hybrid approach becomes the standard, we could see a real shift in how companies communicate — faster, more personalized, but still authentic where it counts.