OpenAI Retires Custom GPTs: Key Insight for Microsoft Copilot Users
OpenAI has announced the retirement of custom GPTs, raising immediate questions about the future of Copilot agents within Microsoft's ecosystem. Here's a detailed analysis of the current situation and what to watch for.
Microsoft's Stance: Agents Are Not Disappearing
Official position: Microsoft is publicly stating that Copilot agents are safe and they are doubling down on agent investment.
Key Microsoft investments:
- Microsoft 365 Copilot – core AI assistant integrating across Office apps
- Copilot Studio – platform for building custom AI assistants, which you can learn to use in our Complete Guide to Copilot Agents in Microsoft 365: Free vs Paid
- Agent 365 – new platform specifically for managing, governing, and securing agents at enterprise scale
Bottom line: No evidence currently suggests Microsoft plans to retire Copilot agents.
The Shift: Skills vs. Agents
Copilot Co-work now supports skills – reusable capabilities that teach the AI how to complete specific tasks or workflows.
What a skill contains:
- Instructions
- Processes
- Templates
- Reference material
- Rules the AI must follow
What an agent contains:
- Knowledge
- Skills
- Tools
- Workflows
- Data
- Actions (can execute work autonomously)
How Microsoft Is Integrating Skills Inside Agents
| Feature | Role | Example | |---------|------|---------| | Skill | Reusable capability | Proposal writing process | | Declarative Agent | Combines skills + knowledge | Marketing assistant with writing skill | | Copilot Studio Agent | Advanced agents with skills | Autonomous process automation |
Current architecture:
- Agent decides what needs to happen
- Skills teach how to complete individual pieces of work
The Speculation: Could Skills Replace Simple Agents?
Historical pattern: Microsoft often adopts patterns from OpenAI/Anthropic a few months later.
Potential future scenario:
- Ordinary Copilot chat uses skills as easily as Co-work does today
- Users may not need separate agents for specific tasks if the main AI can auto-select the right skill
Questions to consider:
- Do you need a dedicated proposal agent if Copilot can automatically use your proposal writing skill?
- Do you need separate marketing, research, writing, or reporting agents if the main AI grabs the right skill on demand?
Why Enterprise Agents Aren't Going Anywhere
Sophisticated agents solve different problems than skills:
- Require their own tools, permissions, and workflows
- Need specific triggers and governance
- Handle complex, autonomous processes
Microsoft's commitment:
- Launched Agent 365 for enterprise-scale agent management
- Continuing investment in Building AI Agents with Microsoft Copilot Studio, autonomous workflows, and agent governance
What to Watch
The interesting question isn't will agents disappear? , it's which jobs still need a dedicated agent once the AI can load skills it needs?
- Short term: Agents thrive alongside skills
- Medium term: Skills may replace simple, instruction-only agents
- Long term: Enterprise-grade agents with custom tools and governance remain essential
Key takeaway: OpenAI's move is a preview of a future direction, not an immediate warning. Microsoft's agent strategy remains strong, but skills will likely reshape how simple AI tasks are handled.
For a step-by-step guide on building your own assistants, see Creating Custom Agents with Microsoft 365 Co-Pilot: A Step-by-Step Guide.
Want more practical AI strategy updates? Subscribe for breakdowns of what AI changes actually mean for your business.
OpenAI has just announced that it's retiring custom GPTs. So, if you've spent time building agents inside
Microsoft Copilot, there's probably one fairly obvious question. Are Copilot agents next? And should we be looking at
what OpenAI has just done as a bit of an early warning sign? Well, based on what Microsoft is saying publicly right now,
the answer is no. In fact, Microsoft seems to be heading in the opposite direction. They're still investing
pretty heavily in agents through Microsoft 365, C-Pilot, Copilot Studio, and Agent 365. And Agent 365 is actually
designed specifically to help organizations manage, govern, and secure agents. So, right now, there really
isn't any evidence that Microsoft is preparing to retire them. But there is one development that I think is worth
keeping an eye on. Copilot Co-work now supports skills and that means you can create your own reusable skills that
tell co-work how you want a particular task or workflow completed for things like your reporting, erocess, proposal
writing, research method, or even the way you create content. And when you look at that, it starts to sound
remarkably similar to the direction Open AI is now heading. But here's where it gets interesting. Microsoft isn't
currently positioning skills as a replacement for agents. They're actually putting skills inside agents. Microsoft
has recently added custom skills as a component of declarative agents. And Copilot Studio agents using the newer
GitHub copilot architecture can also contain skills. So the architecture is starting to look a bit more like this.
The agent decides what needs to happen and then the skills help teach it how to do the individual pieces of work. And I
think that distinction is really important. A skill is basically a reusable capability. So it might contain
things like your instructions, y process, templates, reference material, and the rules you want the AI to follow.
Whereas an agent can potentially bring all of that together. It might have knowledge, skills, tools, workflows,
data, and actions. And depending on how it's built, it can actually go off and do work on your behalf. Now, this next
bit is definitely more speculation. One thing I've noticed over the last few years is that Microsoft quite often
adopts patterns that we first see emerging from companies like OpenAI and Anthropic, sometimes a few months later.
So, I do think OpenAI retiring custom GPTs is worth paying attention to. Not because Microsoft has said agents are
disappearing. They haven't. But because skills could eventually make some of those really simple agents unnecessary.
Think about it this way. Imagine ordinary co-pilot chat eventually being able to use your skills as easily as
co-work does today. Do I really need a separate proposal agent if I can just ask co-pilot to create a proposal and it
automatically uses my proposal writing skill? Do I need separate little agents for marketing, research, writing, or
reporting if the main AI can simply grab the right skill when it needs it? And that to me is where this starts getting
really interesting because that's effectively the question Open AI has just answered with custom GPTs. But I
definitely wouldn't write off agents yet. Microsoft has just launched an entire platform called Agent 365
specifically to manage agents at enterprise scale and they're still investing in agent builder co-pilot
studio autonomous workflows and agent governance. So that's a pretty significant commitment and more
sophisticated agents that need their own purpose tools permissions workflows triggers and governance still solve a
very different problem from a reusable skill. So, are co-pilot agents about to disappear? There's absolutely no
evidence of that today. But could skills gradually replace some of those smaller specialized agents that are really just
a set of instructions wrapped around an AI? I think that's the bit worth watching because for me, the interesting
question isn't really will agents disappear. It's which jobs still need a dedicated agent once the AI itself can
just load the skills it needs. and OpenAI may have just given us the first glimpse of where that could eventually
lead. If you want to hear more important AI updates, practical strategies for adopting AI in your business, and what
these changes could actually mean for the way you work, why not like and subscribe? I'll keep breaking down the
important AI changes, what matters, what doesn't, and what you should actually be paying attention to in your
OpenAI announced the retirement of custom GPTs, prompting speculation about the future of similar AI tools. For Microsoft Copilot users, this raises concerns about whether Copilot agents might face the same fate. However, Microsoft has stated that Copilot agents are safe and it is doubling down on investments in this area.
Microsoft publicly states that Copilot agents are not being retired. The company is investing in agents via Microsoft 365 Copilot, Copilot Studio for building custom assistants, and Agent 365 for enterprise-scale management, governance, and security.
A skill is a reusable capability that teaches the AI how to complete specific tasks by providing instructions, processes, templates, reference material, and rules. An agent, on the other hand, combines knowledge, skills, tools, workflows, data, and actions to execute work autonomously. Agents are more complex and can handle independent processes.
Yes, in the medium term, skills may replace simple, instruction-only agents. If the main AI can automatically select and apply the right skill for a task, users may no longer need separate agents for basic functions like writing proposals or conducting research. However, enterprise-grade agents with specialized tools, permissions, and governance will remain essential.
Agent 365 is a new platform from Microsoft designed specifically for managing, governing, and securing agents at the enterprise scale. Its launch demonstrates Microsoft's commitment to sophisticated agents that require their own tools, permissions, workflows, triggers, and governance for handling complex, autonomous processes.
In the short term, agents will thrive alongside skills. In the medium term, skills may replace simple, instruction-only agents. In the long term, enterprise-grade agents with custom tools and governance will remain essential. The key question is which jobs still need a dedicated agent once the AI can load skills on demand.
OpenAI's retirement of custom GPTs is a preview of a future direction, not an immediate warning for Microsoft users. Microsoft's agent strategy remains strong, but the shift will reshape how simple AI tasks are handled, with skills taking over routine jobs. Businesses should monitor how skills evolve to reduce reliance on dedicated agents for basic tasks.
Keep this summary
Save it to LunaNotes and it becomes a real note in your library — editable, searchable, and ready to turn into flashcards or a diagram. Free to start.
Save to LunaNotesOr summarise for another video.
32K+ people have turned 65K+ videos into notes
“Very useful tool. I'm using it for the last couple weeks and I'm loving it so far. Makes learning on YouTube a great experience.”
This summary and transcript were automatically generated using AI with the Free YouTube transcript summary tool by LunaNotes.
Related summaries
Complete Guide to Copilot Agents in Microsoft 365: Free vs Paid
Master Copilot agents in Microsoft 365, from basic concepts to advanced usage. This comprehensive tutorial covers both free Copilot Chat and paid Microsoft 365 Copilot licenses, showing you how to build custom agents, use out-of-the-box Researcher and Analyst agents, and maximize productivity with grounded knowledge sources.
Build AI Agents with Microsoft Copilot Studio – No Code Guide
Learn how to create custom AI agents using Microsoft Copilot Studio without writing any code. This step-by-step tutorial covers building a customer service bot, connecting it to company documents, automating email responses, and publishing it for your team.
AI Agents Explained: The Simple Framework to Stay Ahead
Stop treating AI as a glorified search engine. This video breaks down the core difference between prompts and agents using the ARR framework, explains the anatomy of an agent with four key workers, and shows you how to leverage autonomy for recurring tasks, avoiding the common pitfalls that sink most implementations.
Creating Custom Agents with Microsoft 365 Co-Pilot: A Step-by-Step Guide
This video provides a comprehensive overview of the Agent Builder experience within Microsoft 365 Co-Pilot, enabling users to create personalized agents without any prior skills. It covers the process of customizing agents for specific business needs, connecting to various data sources, and refining agent functionalities.
Free 1-Hour Microsoft Copilot Course: Pro Tips & AI Agent Creation
Master Microsoft Copilot with this comprehensive free course from Microsoft MVP David Foron. Learn how to prompt effectively, automate Excel tasks, manage Teams meetings, and build your first custom AI agent to save hours each week.
Most viewed summaries
A Comprehensive Guide to Using Stable Diffusion Forge UI
Explore the Stable Diffusion Forge UI, customizable settings, models, and more to enhance your image generation experience.
Kolonyalismo at Imperyalismo: Ang Kasaysayan ng Pagsakop sa Pilipinas
Tuklasin ang kasaysayan ng kolonyalismo at imperyalismo sa Pilipinas sa pamamagitan ni Ferdinand Magellan.
Mastering Inpainting with Stable Diffusion: Fix Mistakes and Enhance Your Images
Learn to fix mistakes and enhance images with Stable Diffusion's inpainting features effectively.
Pamamaraan at Patakarang Kolonyal ng mga Espanyol sa Pilipinas
Tuklasin ang mga pamamaraan at patakaran ng mga Espanyol sa Pilipinas, at ang epekto nito sa mga Pilipino.
How to Install and Configure Forge: A New Stable Diffusion Web UI
Learn to install and configure the new Forge web UI for Stable Diffusion, with tips on models and settings.
Found this summary useful?
Take it with you. One click puts it in your own LunaNotes library.
Save to LunaNotes