AI Won’t Replace Your Team — It’ll Hand Back the Boring 80%
There’s a question underneath AI conversation today. Mostly it’s not said out loud... it just hangs in the air: “If this really works... how many people does my company not need?”
I understand the instinct. But after more than two years of running Copilot engagements, I’ve landed somewhere different, and Microsoft’s research now backs it up. AI is genuinely excellent at the first 80 to 90 percent of a task. It is not good at the last 10 to 20 percent. And that last slice, the judgment, is exactly where the value has always lived.
So the headline isn’t “AI replaces your team.” It’s “AI redraws the last 20%.” Here’s what I mean, and why it should change how you lead.
The 80/20 reality
Watch what Microsoft Copilot is actually good at and a pattern jumps out. For example Copilot is great at: drafting an email, building the first version of the deck, summarizing an hour-long meeting into a page, reconciling invoices, and even writing the first draft of this blog! (blog prompts were my personal knowledge and experience and information from the vast number of calls I have on Copilot every week!)
That’s the 80%. The volume work, the blank-page work, the tedious-but-necessary work.
What it can’t do is the part that comes next:
- Is this right?
- Does it reflect our strategy?
- Would I actually say this to this client?
That’s the 20%, and it’s human.
Microsoft’s 2026 Work Trend Index found that “86% [of people] say they treat AI output as a starting point, not a final answer,” and that they “stay responsible for the thinking.” The report calls it “the new agency equation: as agents take on more of the execution, humans increasingly have more agency more room to direct the work, make the calls, and own the outcomes.” That’s a promotion to your team if you lead it as one.
Where Agents and Skills Fit
Agents do not change the 80/20 reality. They make it more obvious. If Copilot helps with the first draft of a task, agents start to take on the repeatable steps inside a process. They monitor the signal, gather the information, move the work forward, and stop when the process needs human judgment.
Take a sales order process in Dynamics 365 Business Central.
A Sales Order Agent can:
- Monitor a shared inbox
- Read a customer’s request
- Identify the customer
- Find the item
- Check availability
- Prepare a quote or order.
That is the 80%.
The human comes alongside the agent when something needs judgment:
- The customer used vague language
- The item match is uncertain
- The delivery promise feels risky
- The outgoing quote needs approval before it goes to the customer
The same pattern shows up in accounts payable.
A Payables Agent can:
- Read invoices
- Match vendors and accounts
- Prepare purchase invoices for approval
But the human still owns the important calls:
- Is this vendor right?
- Does this invoice look unusual?
- Should this exception be approved, questioned, or escalated?
The agent handles the motion. The person handles the meaning.
A skill is the repeatable capability an agent can use: check inventory, read an invoice, compare a vendor record, draft a customer response, update a case, or prepare a document for approval.
Those skills live inside the 80%. They are the structured, repeatable pieces of work that are painful for humans to do over and over, but still need clear boundaries, good data, and someone accountable for the outcome.
Agents are better understood as process teammates. They take on the repeatable movement of work, while people define the rules, watch the exceptions, improve the process, and make the decisions that carry business risk. The job is not disappearing. It is becoming more supervisory, more strategic, and more accountable.
Human in the Loop is by Design, Not by Accident
Here’s the part I think gets missed: keeping a person in the loop isn’t some awkward workaround because AI isn’t good enough yet. It’s the point. Microsoft has designed these tools assuming that humans are still responsible for the outcome.
Even newer features like Copilot Notebooks are built around that idea. A Notebook lets a person pull together the right files, notes, prompts, and context so Copilot can help with the work but the human is still deciding what matters, what to trust, and what is ready to move forward.
The same pattern shows up in business applications like outlined above. In Dynamics, agents can handle a lot of the routine processing around things like quotes and invoices. But when the system needs a decision, approval, or a closer look, it brings a person back into the process. It even highlights low-confidence information, so someone knows where to slow down and review carefully.
That’s not AI running wild and humans trying to catch up. That’s AI doing the heavy lift, then handing the right moments back to the people who own the decision.
What Actually Changes About the Work
So if it’s not really about headcount, what does change? The shape of the job.
Your best people should not be spending their mornings stuck in first-draft mode anymore. That is not where their highest value is. AI can help create the rough draft, summarize the meeting, organize the notes, compare the options, or pull together a starting point. But that does not mean the human work gets smaller. It means the human work gets more important.
This is the part leaders need to define clearly. In an AI-enabled job, the person’s role shifts from simply producing the first version of the work to shaping, checking, challenging, and owning the final version. They decide what context matters. They recognize when something sounds right but is missing the point. They catch the nuance, risk, customer impact, or business implication that the tool does not fully understand. They ask the better follow-up question. They decide what is ready to move forward and what still needs more thought.
That is the new value of the job.
The person who used to write the report now:
- Reviews it
- Sharpens it
- Pressure-tests the assumptions
- Stands behind the final recommendation
The person who used to compile the notes now identifies the decision, the risk, and the action.
The person who used to start from a blank page now starts from a better draft and spends more of their time applying judgment. Same person, but the work moves up a level.
I have seen what happens when that human step gets skipped. I once watched a leader read an AI-generated summary to a room almost word for word, without doing a real pass over it first. You could feel the trust drop-in real time. Not because the tool was bad, but because nobody owned the output. Nobody paused long enough to ask, “Is this right? Is this useful? Is this what this group actually needs to hear?” Note: I've also seen the reverse to be true... a truly magical speech by a leader whose first draft was written by Copilot!
That is the part we cannot automate away. Speed is great, but speed without judgment just helps us make bigger mistakes faster. The win is not replacing the person. The win is giving that person a better starting point so they can spend more time on the work that actually requires them.
How to Lead AI strategy (not just deploy it)
The organizations getting real, durable value out of AI aren’t the ones treating it like autopilot. They’re the ones treating it like a very capable, very fast colleague who still needs a human editor. A few things that separate them:
- Make review a step. Build in human checks before things go out into the workflow — especially for anything client-facing or financial.
- Name who owns oversight. As we put it on our own site, teams need clear guidance on “human review,” “accountability,” and “who owns oversight and governance.” Ambiguity is where things go wrong.
- Invest in the people, not just the licenses. The value doesn’t come from the purchase order. As I’ve written before, adoption is a practice, not an event. Teach your people where their judgment adds the most, and the 20% becomes your competitive edge instead of your risk.
The Bottom Line for AI vs. People
Here’s the reframe I’d offer any leader staring at this technology and quietly doing headcount math. The winners won’t be the companies that used AI to replace their people. They’ll be the ones that used it to hand their people the boring 80% back — so the humans can spend their time, energy, and judgment on the 20% that was the actual job all along.
AI is a phenomenal way to get to a first draft. Your team is still what makes it true.
If you're exploring how to put AI to work while keeping people at the center of the process, we'd be happy to talk. You can also learn more about our Flight Plan framework for building a thoughtful approach to Copilot adoption.
Sources
All references are Microsoft or Stoneridge Software public sources.
- Microsoft — Responsible AI principles and approach — https://www.microsoft.com/en-us/ai/principles-and-approach
- Microsoft WorkLab — 2026 Work Trend Index (agents and human agency) — https://www.microsoft.com/en-us/worklab/work-trend-index/agents-human-agency-and-the-opportunity-for-every-organization
- Microsoft — Copilot Terms of Use — https://www.microsoft.com/en-us/microsoft-copilot/for-individuals/termsofuse
- Microsoft 365 Copilot Blog — What’s new in Microsoft 365 Copilot, June 2026 (deep citations) — https://techcommunity.microsoft.com/blog/Microsoft365CopilotBlog/what%E2%80%99s-new-in-microsoft-365-copilot--june-2026/4529572
- Microsoft Learn — Supervise agent activities in Business Central — https://learn.microsoft.com/en-us/dynamics365/business-central/supervise-agent-tasks
- Stoneridge Software — AI Strategy and Adoption — https://stoneridgesoftware.com/ai-strategy-and-adoption/
- Stoneridge Software — From License Purchase to Landed Value: The Flight Plan Playbook for Copilot Adoption — https://stoneridgesoftware.com/from-license-purchase-to-landed-value-the-flight-plan-playbook-for-copilot-adoption/
Under the terms of this license, you are authorized to share and redistribute the content across various mediums, subject to adherence to the specified conditions: you must provide proper attribution to Stoneridge as the original creator in a manner that does not imply their endorsement of your use, the material is to be utilized solely for non-commercial purposes, and alterations, modifications, or derivative works based on the original material are strictly prohibited.
Responsibility rests with the licensee to ensure that their use of the material does not violate any other rights.