Dear Friends,
Some weeks in corporate life feel like a tug-of-war between what we want technology to do for us, and what it’s actually allowed to do.
If you’ve used Copilot, you probably know the feeling. It’s safe, secure, and beautifully aligned with all the policies, IT frameworks, risk committees, and data boundaries we need in a corporate environment. It lives inside Teams, SharePoint, Outlook, Excel—everything we touch every day. Microsoft states clearly that Copilot does not train on your corporate prompts or content.
All of this is reassuring, but also limiting. When you’re deep inside a live project, searching for clarity, trying to interpret models, or piecing together business decisions from meeting minutes, you can feel the boundaries of the tool very quickly. It knows our organisation, but it doesn’t always understand our questions. It follows rules, but sometimes misses meaning. And when the pressure is on—tight deadlines, high expectations, credibility at stake—you start wishing the AI could think a little deeper, reason a little better, and maybe even read your mind.
And then—just as I was feeling this tension—OpenAI dropped a new feature called Company Knowledge, and I found a striking demo of how it works with Box AI. In that YouTube video, the system was shown “searching everything in our shared folders and providing deep, contextual answers across Slack, Drive and SharePoint”.
On the surface, it looks like the thing we’ve always wanted.
It promises the best of both worlds: the intelligence of GPT-tier models combined with your organisation’s internal knowledge pulled in from multiple sources. OpenAI+1 If this works as advertised, it could be transformational: a genuinely smart AI that understands both the internet and your company’s data. Every analyst, finance manager, transformation lead, project team would benefit.
I can already imagine how different my project would feel if I could ask:
- “What funding decision did we agree in Q2 for this project?”
- “What documents have referenced our debt drawdown since June?”
- “What assumptions were changed in the last reforecast, and which meetings discussed them?”
It does sound like magic.
But as always, corporate life gently brings us back to reality. Because the question is never just “How smart is the tool?” It’s also “Can we safely use it?” Before some companies can adopt OpenAI’s Company Knowledge, serious considerations remain around data residency, access control, audit logs, retention policies and risk. Governance must be watertight—not just for compliance, but to protect clients, stakeholders and our own reputations.
That’s why Copilot became the default assistant for so many companies: it’s already inside the ecosystem, fits existing permission models, and is built for enterprise trust—even if its reasoning is less impressive than the newest model. OpenAI themselves note that enterprise control, permissions and encryption remain core.
So I’m watching the evolution with curiosity, hope, and a sense of healthy caution. The future is coming fast, and tools like Company Knowledge show us what’s possible. But real adoption will take time. Governance always moves slower than innovation—and maybe that’s how it should be.
If there’s one takeaway I’m landing on this week, it is this:
The most valuable skill for professionals today is not knowing how to use AI — it’s knowing how to think about AI.
How to balance curiosity with responsibility.
How to welcome intelligence without compromising integrity.
How to explore new tools without exposing our organisations.
How to stay adaptable even when technology changes faster than we can read the user guide.
I still believe the perfect AI tool will eventually arrive—one that is as safe as Copilot and as smart as OpenAI. Until then, I’ll continue navigating my project the old-fashioned way: one model, one meeting, one minute at a time, supported by tools that help just enough—but not too much.
If you’ve also been navigating these new tools, or wondering how AI will reshape our work, I’d love to hear your thoughts. We’re all learning this together, week after week.
Until next time —
Lydia
P.S. If this letter found you at just the right moment, I’d love to hear about it. Join my weekly letter list and let’s figure it out together — one AI-shaped step at a time. Join the weekly letter list.
☕💌 If you’d like to fuel my next cup of coffee and keep this journey going, you can:
- 💳 Buy me a coffee via Stripe
- 🅿️ Buy me a coffee via PayPal
Your support keeps the ideas flowing and the coffee brewing ✨