Microsoft Just Scored a $9.7 Billion Pentagon
Deal — And That's Just the Start
Published May 2026 | Finance & Tech | 10 min read
I'll be real with you — when I first saw this headline, I almost scrolled past it.
"Microsoft wins $9.7 billion Pentagon deal." Cool. Big number. Moving on.
But something made me stop and actually read the whole thing. And the more I dug into it, the more I realized this story is way bigger than just one contract. It's really about what Microsoft is turning into right now — and honestly, if you follow the stock market or just care about where tech is heading, this is one of those moments worth paying attention to.
So let me walk you through everything — the Pentagon deal, the AI stuff, the financials, all of it. No jargon, no fluff. Just what's actually happening and why it matters.
What Is This Pentagon Deal, and Why Should You Care?
Okay, so here's the deal in plain English.
The Department of Defense just signed a five-year, $9.69 billion contract with Microsoft called the Core Enterprise Technology Agreement. And before you picture the Pentagon buying some fancy new military AI supercomputer — that's not what this is.
This is basically the government finally getting its act together.
What was happening before is kind of embarrassing, honestly. The Army was buying Microsoft software through one contract. The Navy through another. Intelligence agencies through their own separate deals. The Coast Guard doing the same thing. Everyone buying the same stuff, separately, paying whatever they were quoted, with zero coordination.
It's like if your whole family each drove to a different grocery store and bought the same cereal at full price, when you could've just sent one person with a big cart and gotten a bulk discount.
This new deal fixes all of that. Every branch of the military, every intelligence agency, the Coast Guard — they're all now under one single Microsoft contract. One negotiation. Better pricing. Way less paperwork.
But Wait — What Software Are We Talking About?
The stuff these agencies already use every day. Microsoft 365 — that's Word, Excel, Outlook, Teams, the whole suite. Cloud subscriptions. Other software tools that are already running across the entire defense ecosystem.
Nothing new is being installed. Nobody's getting some experimental tech. This is literally just the government saying "let's stop being wasteful and buy this stuff together like adults."
And that actually tells you something important — Microsoft's software is so deeply embedded in how the U.S. military operates that the solution to cutting costs wasn't to switch to something else. It was to just buy more Microsoft, but smarter.
Why Doesn't the Government Just Switch to Google or Something Else?
Great question, and the short answer is: it's not that simple.
Microsoft has spent years — honestly, decades — meeting the insane security requirements that government agencies demand. Things like FedRAMP authorization and DoD Impact Level certifications. These aren't boxes you check in a weekend. They require massive investment, constant auditing, and a long track record of compliance.
Google and others are working on this too, but Microsoft has a serious head start. And when you're dealing with sensitive military and intelligence data, the last thing any agency wants to do is risk a switch to something unproven just to save a few bucks.
What Does This Actually Mean for Microsoft's Business?
Here's my honest read on this — and I want to be straight with you rather than just hype it up.
This deal isn't going to make Microsoft's stock double overnight. Analysts are pretty clear that it's not a sudden growth shock. You're not going to open your brokerage app tomorrow and see MSFT up 25%.
What it does is lock in five years of rock-solid, predictable income from one of the biggest technology buyers on the planet. And in the tech business, predictable revenue is genuinely underrated. Most companies are out there fighting quarter to quarter for every dollar. Microsoft just secured nearly $10 billion over five years from a customer that literally cannot easily leave.
That's a great position to be in.
Think about it from a business perspective — once the U.S. military is running on your software, the cost and hassle of switching to a competitor is so enormous that it almost never happens. You'd have to retrain hundreds of thousands of people, migrate years worth of classified data, re-certify everything from a security standpoint. It's a nightmare no government IT department wants to touch.
So Microsoft isn't just winning a contract. It's deepening a relationship that was already basically impossible to break.
Microsoft Is Building Its Own AI — And the Stock Already Jumped 3%
Now here's where this story gets really interesting, and where I think the real opportunity is.
For the past couple of years, Microsoft's AI features — the Copilot stuff you see everywhere in Word, Teams, Outlook — have been powered mostly by OpenAI's models. Microsoft invested billions in OpenAI and got access to their technology at no cost through a deal that runs until 2032.
That sounds great. And it was — for a while.
But here's the thing nobody talks about enough: that deal ends. And when it does, if Microsoft has to start paying full market rate for OpenAI's models, the economics of running Copilot get a lot uglier really fast. Your margins shrink, your costs go up, and suddenly the AI products that are supposed to be your growth engine start eating into your profits.
So Microsoft started working on something quietly. Their own AI models.
What Exactly Is Microsoft Building?
At the Build developer conference in San Francisco, Microsoft unveiled a whole new lineup of AI models built entirely in-house. We're talking about:
- A coding model built specifically to make GitHub Copilot better — and more competitive with tools like Cursor and Claude Code that have been nipping at its heels
- Reasoning models for handling complex, multi-step tasks
- Speech and transcription models to make Teams and voice tools smarter
- Image generation models for creative and business workflows
The guy running all of this is Mustafa Suleyman — Microsoft's Chief AI Officer and one of the co-founders of DeepMind. This isn't some junior product team tinkering in a corner. These are serious people building serious technology.
Here's the Part Nobody Saw Coming
There's actually a reason Microsoft couldn't do this before. Their original deal with OpenAI had restrictions that prevented Microsoft's internal team from training their own top-tier models. Basically, they agreed not to directly compete with the technology they were paying to use.
That restriction got renegotiated in April 2026. And the moment it did, Microsoft went full speed ahead on building their own.
The market reacted immediately. MSFT stock jumped 3% in a single day when news of these in-house models broke. For a company worth over $3 trillion, a 3% move in one session is a big deal. It tells you investors had been waiting for exactly this moment.
The plan is to use these new models as cheaper alternatives to OpenAI and Anthropic's offerings for everyday tasks. Same Copilot experience for the user sitting in front of their laptop — but way better profit margins on the back end for Microsoft. That's the kind of business move that compounds quietly and pays off massively over time.
Copilot Is Taking Off Faster Than Almost Anyone Predicted
Let me share some numbers with you — and I promise to keep them human.
Microsoft just released its fiscal third-quarter results for the period ending March 31, 2026, and the Copilot numbers genuinely surprised people. Here's what stood out:
Microsoft 365 Commercial cloud revenue was up 19% year over year — that beat even Microsoft's own internal expectations
- The whole Productivity and Business Processes segment (think Office, Teams, Copilot) hit $35 billion in revenue, up 17%
- Over 20 million people are now paying for Copilot seats
- Seat additions grew 250% year over year — the fastest rate since Copilot launched
- The number of companies with more than 50,000 Copilot seats? It quadrupled
That last one is the one I keep coming back to. These aren't small businesses dipping a toe in. These are massive enterprises going all-in on Copilot at scale. When a company with 50,000+ employees rolls out Copilot company-wide, that's a serious, long-term revenue commitment.
And here's something that doesn't make headlines but should — efficiency improvements in the M365 cloud actually helped gross margins tick up modestly, even while Microsoft is spending heavily on AI infrastructure. That means Copilot isn't just driving revenue — it's doing it without wrecking the economics. That's rare at this stage of a new product's growth.
What's Pushing People to Upgrade?
Two things mostly. First, companies are moving up to higher-priced Microsoft 365 tiers — the E5 plan and the Copilot bundles — which means more revenue per employee, not just more employees using the product. Second, the new Microsoft 365 E7 "Frontier Suite" and Agent 365 launched May 1, 2026, which gives enterprises even more to upgrade into.
There's also a commercial pricing update coming July 1, 2026. So if you're a CFO looking at your Microsoft bill, it's probably going up. And if you're an investor, that means ARPU — average revenue per user — is likely heading higher.
Okay But Is Everything Perfect?
No, and I'd be doing you a disservice if I pretended it was.
Microsoft 365 Commercial products revenue — the non-cloud stuff, like Office 2024 licenses — grew just 1%, and actually dropped 3% when you adjust for currency. The initial wave of Office 2024 purchases has normalized, which was expected but still a drag.
Paid commercial seat growth came in at 6%, which sounds decent until you realize most of that was small businesses and frontline workers rather than big enterprise clients.
On the bookings side — commercial bookings grew 7% excluding OpenAI on solid annuity contracts, which is healthy. But Dynamics 365 had weaker renewals as customers shift from paying per seat to a mixed model of seats plus consumption-based pricing. That transition is real, and it's creating some messiness in the numbers quarter to quarter.
Microsoft's own Q4 guidance calls for M365 Commercial cloud growth of 15-16% in constant currency (13-14% on a reported basis), with sequentially higher net paid seat adds — meaning they expect Copilot to keep picking up steam going into the second half of the year. Overall Productivity segment revenue is projected at $37 to $37.3 billion.
So yeah — strong overall, with a few rough edges. That's normal for a business this size in the middle of a big product transition.
Google and Salesforce Are Coming — But Microsoft Has the Home Field Advantage
Let me be honest: Microsoft isn't the only one in this race.
Google has been building Gemini directly into Google Workspace, and they're not messing around. They're reporting more than 8 million Gemini Enterprise paid seats across about 2,800 companies. And their paid monthly active users jumped 40% in a single quarter. That's real growth — not hype, not press releases, actual paying customers.
Salesforce is pushing Agentforce hard. Their approach is a little different — they're selling it through consumption-based Flex Credits alongside traditional per-user add-ons. The pitch is mainly around customer service: deflecting support tickets automatically and automating business workflows. It's a slightly different angle than Copilot, but the overlap is growing.
So Who's Actually Winning?
Honestly, it's still too early to say. All three companies are genuinely growing, and each has a real angle.
But here's what I keep coming back to with Microsoft specifically. Google and Salesforce have to convince companies to add something new to their workflow. Microsoft already lives inside the office. Your employees are already in Teams, already in Outlook, already staring at a Word document. Adding Copilot to that is a completely different conversation than asking IT to approve a brand new vendor.
You're not selling something new. You're activating something people technically already have access to. That's a huge advantage in enterprise sales, and I don't think it gets talked about enough.
The Numbers Investors Actually Care About
Alright, let's get into the financials properly — because this is the stuff that matters if you're watching MSFT.
The 1-Year Price Target: That $870 Number Is Real
I want to spend a minute on this because I think it deserves more than a one-liner.
The analyst consensus target for MSFT over the next 12 months is $559.33. That's the average across all the analysts covering the stock — and it already implies meaningful upside from where shares are trading today.
But the maximum analyst target is $870.
That's the number from the most bullish analyst covering Microsoft right now. And that target isn't just wishful thinking — it's based on a specific scenario where Copilot adoption keeps accelerating at its current pace, Azure keeps taking cloud market share, the in-house AI models gain real commercial traction, and the government contracts provide a steady, growing baseline. If all of those things line up the way the bulls expect — $870 in a year isn't crazy.
Now, will it hit $870? Nobody knows. The most conservative analyst has a target of just $400, which tells you there are smart people who think the AI story is mostly priced in already or that macro risks could weigh on the stock. That gap from $400 to $870 shows you just how much genuine disagreement there is right now.
What I'd take from this: the ceiling is high if things go right. Just don't bet your retirement on the most optimistic scenario.
EPS Keeps Beating Expectations
Last quarter, Microsoft's earnings per share came in at $4.27 against estimates of $4.06. That's a solid beat — and for a company this size, consistently beating EPS estimates is a sign the business is running well even while investing heavily in new things.
Next quarter EPS is expected around $4.24. Slightly lower than last quarter, which is fine. What matters more is whether Microsoft beats again — and based on recent history, there's a decent chance they do.
Revenue: Nearly $88 Billion in a Single Quarter
Next quarter revenue is projected at roughly $87.61 billion. That is an almost absurd amount of money to generate in 90 days. It puts in perspective just how massive and diversified this business actually is.
What 60 Analysts Are Saying
Sixty analysts have weighed in on MSFT stock over the past three months. After looking at all their ratings — the optimistic ones, the cautious ones, everything in between — the consensus came out as Strong Buy.
That's meaningful. These aren't random people on Twitter. These are professionals who study Microsoft's financials full time. When the majority of them land on Strong Buy, it's worth noting.
Again — not a guarantee of anything. The market does what the market does. But it's a useful data point.
The Recent Stock Drop Is Frustrating — Here's How to Think About It
MSFT shares have fallen about 15.3% over the past six months. That's roughly in line with the broader software sector, which dropped about the same amount. But it's a painful contrast to the wider tech sector, which actually grew around 18.6% in the same period.
So yeah — if you've been holding Microsoft stock lately, it's been a rough ride while watching other tech stocks run.
Here's my honest take on it though. A six-month stock chart doesn't tell you whether a business is good or bad. The revenue is growing. EPS is beating estimates. Copilot is scaling faster than expected. The Pentagon deal is signed. None of that changed because the stock pulled back.
Sometimes strong businesses go through stretches where the stock just doesn't reflect what's actually happening inside the company. Whether that's true here, I can't say for certain. But I wouldn't write off the business based on six months of stock performance.
Microsoft Is Building Enough Data Centers to Power a Small Country
I want to talk about something that most people completely overlook when they think about Microsoft's AI strategy — the physical infrastructure behind it all.
Here's a number that blew my mind when I first read it: Microsoft is expanding its data center capacity from about 5 gigawatts in 2024 to roughly 20 gigawatts by 2028. That's four times the capacity in four years according to Morgan Stanley.
Now, gigawatts is an electricity term, not a computing term — but that's kind of the point. AI is an energy business as much as a software business. Running these large AI models, answering millions of Copilot queries every day, processing data at this scale — it burns through electricity at a rate that's genuinely staggering.
If you don't have the physical infrastructure, you literally can't scale. It doesn't matter how good your software is. You hit a wall.
By going from 5 to 20 gigawatts, Microsoft is essentially betting big that demand for AI computing is going to keep growing for years. And they're making sure they have the capacity to handle it — and to sell that capacity to other companies who need it.
The Anthropic Piece of This
Here's something interesting about the Anthropic partnership. Microsoft and Anthropic — the company behind the Claude AI — have deepened their relationship, with Anthropic committing to run a lot of its computing workloads on Azure.
Some people find this confusing. Microsoft is partnered with OpenAI AND Anthropic? Aren't those kind of competitors?
Yes, kind of. But from Microsoft's perspective, it doesn't really matter. When Anthropic runs on Azure, Microsoft gets paid for the compute. It's like owning a highway and charging everyone who drives on it — whether they're going north or south, you're collecting the toll either way.
Having major AI labs running on Azure also makes the platform look more attractive to every other AI startup and enterprise company that wants to be where the serious AI work is happening. Analysts expect this to be a meaningful source of revenue growth for Microsoft through 2030.
The Quiet Safety Update That Actually Says a Lot
This one barely got any coverage but I think it's worth a minute.
Microsoft quietly updated its reporting system for Non-Consensual Intimate Images — NCII — the kind of harmful content that gets created and shared without someone's consent. They made it easier to report, added clearer categories for different types of harm, and specifically created a way to distinguish between real photos and AI-generated images.
Why am I bringing this up in a business blog?
Because as Microsoft rolls out image generation tools and more powerful AI features, they're also taking on more responsibility for what those tools can be misused for. Updating this system is a sign they're thinking about that — before a PR disaster forces them to, not after.
It's a small thing. But companies that handle the hard stuff quietly, without being forced to, tend to build the kind of institutional trust that matters when it's time to sign big government contracts or land enterprise clients who are worried about AI risk. Trust is worth a lot of money in this industry, even when it's hard to put a number on it.
My Honest Take: Should You Be Watching Microsoft Right Now?
Yes. But let me be clear about what I mean by that.
Microsoft isn't a trade. It's not a "buy this week and check back in a month" kind of situation. It's a company that's been slowly, quietly building an almost unbreakable position in enterprise software for decades — and is now using that foundation to push into AI, government infrastructure, and cloud computing all at the same time.
The Pentagon deal alone won't move the needle dramatically. But add that to 20 million Copilot users growing at 250% per year, a $870 one-year price target from the most bullish analysts, a 20 gigawatt data center expansion, and in-house AI models that just sent the stock up 3% in a single day — and you've got a company firing on a lot of cylinders simultaneously.
Is the stock going to go straight up from here? Probably not. It's been down 15% over six months. The Dynamics 365 numbers are soft. The bookings model is in transition. There are real things to watch.
But five years from now? A company that owns enterprise software, AI infrastructure, major government contracts, and its own AI models is going to be in a very tough position to compete with.
Quick Answers to Questions You're Probably Asking
Is this $9.7 billion new money or existing spending? Mostly existing. The government was already paying for Microsoft software — just through a dozen different messy contracts. This deal consolidates everything. Don't expect a sudden revenue spike.
Will Microsoft's own AI models replace OpenAI? Not completely, at least not soon. They still have free access to OpenAI's tech until 2032. The in-house models are positioned as cheaper alternatives for simpler tasks — and a backup plan for when that deal changes.
What numbers should I track going forward? Copilot paid seat growth, Azure revenue, and how the July 2026 pricing update flows through to average revenue per user. Those three things will tell you more than any headline.
What does this mean if I'm just a regular Microsoft 365 user? Expect more AI features showing up in the apps you already use. And expect your company's Microsoft bill to creep up — especially after the July 2026 pricing update kicks in.
Is MSFT a good investment? That's really not for me to say — your situation is different from mine, and I'm not a financial advisor. What I can say is the business fundamentals look solid, 60 analysts say Strong Buy, and there's a credible path to $870 if things break right. Do your own homework and talk to someone who knows your full financial picture.
Let's Wrap This Up
When I first saw that Pentagon headline, I almost didn't write about it.
But it turned out to be the thread that pulled open a much bigger story — about a company that's quietly securing its position in government, in enterprise, in AI, and in the infrastructure that all of it runs on.
The $9.7 billion deal is just one piece. The Copilot growth is another. The in-house AI models. The 20 gigawatt data center buildout. The Anthropic partnership. The 3% stock jump on a single day. The $870 analyst target sitting out there.
None of these things on their own are the story. Together, they're a picture of a company that's trying to be everywhere at once — and mostly succeeding.
Whether the stock reflects that in the next six months is genuinely hard to predict. But the direction Microsoft is moving? That part feels pretty clear.
Watch the July pricing update. Watch the next earnings call. Watch those Copilot numbers. They'll tell you everything you need to know.
Disclaimer:
Everything in this post is for informational purposes only and is not financial or investment advice. Always do your own research and speak with a licensed financial advisor before making investment decisions. Data and analyst targets referenced reflect information available at the time of writing and may have changed.
If this was helpful, share it with someone who's been keeping an eye on Microsoft or the AI space. And drop a comment if you've got questions or a different take — I genuinely read them all.
Post a Comment