Microsoft just did what seemed impossible: they made Notepad controversial.
The 40-year-old text editor, beloved for its simplicity, now requires a Microsoft 365 subscription to access its new AI-powered "Rewrite" feature. Users discovered this when signing into Notepad prompted them to purchase a $7/month plan. The internet responded predictably—with outrage.
But here's what founders should actually pay attention to: this isn't about Notepad. It's about the new economics of AI features and what they mean for every product you're building.
The Real Story: AI as Paywall Engine
Microsoft's move reveals a fundamental shift. Previously, software companies monetized through features, seats, or usage. AI changes the equation because it comes with real marginal costs—every query hits compute. Microsoft isn't being greedy; they're solving a math problem.
When you add AI to a free product, you inherit a cost structure that demands monetization. Microsoft chose subscriptions. But founders have choices.
Three Models for AI Feature Economics
The Microsoft Model: Subscription Gating
Lock AI features behind existing subscription tiers. Works if you have a subscription business already. Forces users into a recurring relationship even if they only wanted the AI occasionally.
The Metered Model: Pay Per Use
Charge per AI call or per "credit." GitHub Copilot does this with premium requests. Transparent but creates friction—users hesitate before each invocation, reducing engagement.
The Loss Leader Model: AI as Acquisition
Offer AI free to drive product adoption, monetizing elsewhere. Notion does this partially, betting that AI-powered users convert to paid plans at higher rates. High risk if your conversion funnel isn't dialed.
The Founder Calculation
Here's the math you need to run. A typical Claude API call costs $0.01-0.05 for input/output tokens. If your free user invokes AI 100 times monthly, that's $1-5/month in hard costs per user. At scale, this destroys unit economics.
Now consider: what's your free user worth? If they convert at 5% with a $20/month price point, your expected value per free user is $1/month. You're underwater on AI costs alone.
Microsoft's solution is aggressive but economically sound. They're saying: we won't subsidize AI usage. If you want it, you pay for it—or you pay for the bundle where it's included.
The Trust Problem Nobody's Discussing
Here's the bigger issue for founders: Microsoft is training users to expect AI features to be premium. This creates friction for every product that wants to use AI as a differentiator in their free tier.
When Microsoft gates Notepad's Rewrite, when Adobe gates Firefly, when Canva gates Magic Design—the market learns that AI costs money. Your free AI feature now gets compared to their paid one. Users assume yours is inferior, or that you'll eventually paywall it too.
This creates a strategic choice: do you compete by being generous (burning cash), or do you match the market's expectations and monetize immediately?
What Smart Founders Are Doing
The founders getting this right are building "AI-native" pricing from day one. Not adding AI to existing tiers, but designing tiers around AI consumption.
Linear offers Cycles (AI-powered sprint planning) as part of their standard plan because AI is core to the experience—not a bolt-on. Notion charges more for AI because it's genuinely premium functionality that requires infrastructure.
The worst position is where Microsoft landed with Notepad: AI as a visible tax on what used to be free. Even if the economics are sound, the optics are terrible.
The Takeaway
Microsoft's Notepad paywall isn't a cautionary tale about greed. It's a preview of how every software company will struggle with AI economics.
If you're building with AI, design your pricing model before you ship the feature. The alternative is retrofitting monetization onto a product where users already expect it for free—and that's a much harder problem than the engineering.
Notepad survived 40 years by being simple. AI killed that simplicity in one update.
