I'm excited and honored to bring our Consumer and Commercial Copilot teams together into one org. When Mustafa Suleyman and I first started talking about this, we kept coming back to the same belief: the best AI product in the world has to be built as one. One team, one product, one experience. Microsoft has everything it needs to make Copilot indispensable for every user. That's the bar, and I can’t wait to continue to drive this work forward with this incredible team.
What does this mean for the frequency and quality of your shitposting going forward?
The "one team, one product, one experience" framing is doing a lot of strategic work here. Consumer and commercial AI assistants have historically followed very different product development logic — consumer products optimize for engagement, discoverability, and delight; commercial products optimize for reliability, auditability, and workflow integration. Unifying the teams doesn't automatically unify those objectives, and the harder question is whether the product architecture can serve both sets of requirements without compromising on either.The case for consolidation is that the distinction between consumer and commercial users is increasingly artificial. Knowledge workers use the same tools at work and at home, expect the same quality of experience in both contexts, and have been trained by consumer AI to have expectations that enterprise software has historically not met. Copilot unifying the experience layer makes sense given that the underlying user is often the same person.The implementation risk is real though: consumer Copilot and commercial Copilot have likely accumulated different technical debt, different feature sets, and different trust models. Org consolidation is the easy part — the harder work is making the product feel like one thing rather than two products sharing a name.
Jacob Andreou congrats on the new promotion. As a user of Copilot just some friendly feedback as it relates to building presentations with a fully built out prompt ranking Copilot against Claude, ChatGPT, Beautiful AI, Gamma, Gemini, Canva and Copilot again from a presentation output creation is a very distant 7. In my view (again I'm neutral here) Claude's presentation output is current #1 in my judgment. These are all the paid versions as I pay for them all on my own. (So Apples to Apples so to speak.) Copilot should clearly be number one in this category (hence this feedback) given Microsoft invented Powerpoint to back in the day so would love to see you as the leader here. You clearly should be. Again big Microsoft fan here, hopefully this constructive feedback is welcomed.
Jacob Andreou Interesting direction—but the current reality looks a bit different. Recent data from Recon Analytics shows that primary usage of Copilot among paid users fell from nearly 20% to just over 10% in six months. In some enterprises, only a small fraction of purchased seats are actively used. The issue doesn’t seem to be fragmentation alone. It’s product-market fit: reliability, clear business value, and workflow integration. Many users still report inaccurate outputs (e.g., broken links, non-existent documents) and struggle to justify the $30/user/month cost—especially when alternatives like ChatGPT or Gemini are improving quickly and seeing increased adoption. Unifying the experience is a good goal—but making it indispensable will depend on solving these fundamentals first.
I'm hoping for good things from a consolidated consumer + business Copilot from Microsoft. I actually made a blog post about this exact topic just over a week before this announcement. https://www.externalprocessor.com/2026/03/why-microsoft-copilot-for-consumers.html
This makes a lot of sense. Users don’t separate their lives into “consumer” and “commercial” they just want one seamless, reliable experience. Bringing everything under one org feels like a strong step toward building something truly indispensable. Excited to see how this unified approach shapes the next phase of Copilot.
"One team, one product, one experience." That framing works at the product layer. The word doing the most work is "every." Indispensable for every user means the same Copilot surface serves a teenager managing homework and a compliance officer at a federally regulated bank. Same product. Fundamentally different accountability requirements. Consumer AI design optimizes for engagement. Enterprise AI design in regulated environments optimizes for auditability and intent traceability. Those are not the same engineering problem and they are not the same governance problem. Building as one is the right call for product experience. The Intent Architecture underneath has to account for the fact that "every user" is not one accountability profile. That is the hard part of what Jacob just signed up for.
Copilot need to get one thing done great instead get everything done ok. We have windows, office, outlook, hotmail, team, etc. let copilot be smart and compliances for personal and enterprise access and get people more productive and creative at work and home. One Copilot, One Agent, One Dream.
Congrats Jacob on joining the Microsoft AI division of the company and I hope you will work hard on the unification of Microsoft Copilot for Commercial Enterprises and Consumers together with other executives in the company moving forward. I hope you will do your best work in strengthening the company’s copilot products and staying competitive with the competitors in the future of AI and the collaboration. This is a great opportunity for leaning and growth mindset intertwine and hopefully this will be helpful to me and to others who are looking forward for a single good Copilot product. So, welcome to Microsoft!😁😁😁😁😁😁😁😁😁
Microsoft AI•2K followers
2whttps://blogs.microsoft.com/blog/2026/03/17/announcing-copilot-leadership-update/