A New Kind of Price Hike: From Seat Licences to Running Meters
Microsoft 365 price hike controversies and the move toward usage-based Copilot billing highlight a shift from predictable subscription fees to metered enterprise AI costs, raising questions about software pricing transparency, bill shock and how clearly vendors must disclose changes to ordinary users and business buyers. That shift is not just about higher numbers on invoices; it is about whether customers are given enough information to understand what they are buying, how much it may end up costing and how their freedom to choose between old and new plans is affected by the way those changes are communicated. GitHub ended its predictable Copilot billing model on June 1 when it moved from a premium request subscription to GitHub AI Credits tied to token consumption. Two weeks later, Microsoft applied similar logic to office work, making Copilot Cowork generally available on June 16 and shifting it to usage-based pricing for enterprises. These changes redefine enterprise software costs, yet the real friction is emerging not from the economics alone but from how those economics are explained to customers.
Italy’s Watchdog Puts Microsoft’s Communications Under the Microscope
Regulators are no longer content to watch from the sidelines while vendors quietly rework their pricing. Italy’s competition authority, the AGCM, launched a regulatory investigation into Microsoft to examine how it informed customers about a Microsoft 365 price hike. This is not a dispute over whether AI features should cost money; it is a dispute over whether Microsoft gave users a fair chance to understand and resist the increase. According to the watchdog, Microsoft may have communicated the price changes and integration of Copilot and Designer AI services in a “fragmented manner,” without making it sufficiently clear that the subscription had changed in scope as well as in cost. Consumers were reportedly shifted to a new, more expensive plan by default unless they actively withdrew. In the AGCM’s view, this deprived them of the information needed to decide whether to renew, raising concerns that the approach “may also constitute an aggressive practice” that restricted users’ freedom of choice.
Bill Shock: When Enterprise AI Starts to Feel Like a Cloud Meter
The regulatory investigation Microsoft faces is happening against a backdrop of mounting bill shock around enterprise AI. Under the old Copilot model, a quick prompt and a multi‑hour autonomous session could cost Microsoft roughly the same while users enjoyed a flat subscription. Once people stopped asking AI for small suggestions and started handing it full tasks, that economic balance collapsed. When GitHub moved Copilot to AI Credits, some users quickly discovered how fast usage could snowball. Business Insider reported that one Reddit user, after the change, projected an USD 847 (approx. RM3,950) bill after previously paying USD 39 (approx. RM180) a month for Copilot Pro+. That extreme case is not representative of every developer, but it shows how opaque metered models can feel when customers have little guidance on what typical workloads cost. If Cowork can edit spreadsheets, compare file versions, schedule meetings and draft messages across a company, finance teams need clarity on which tasks quietly turn a familiar subscription into open‑ended compute spend.
From Bundling Fights to Pricing Transparency: A Pattern Emerges
The Italian probe is not an isolated skirmish; it is the latest chapter in a longer story about Microsoft’s pricing power. Last year, Microsoft ended a multi‑year investigation in the EU by changing how it provided 365, after concerns that bundling its products restricted competition in cloud‑based communication and collaboration tools. The company agreed to unbundle services to address those findings. Today’s focus is narrower but more pointed: not whether a bundle exists, but whether customers understand what they are being moved into and at what cost. This is why treating Microsoft’s moves as mere “pricing tweaks” misses the bigger issue. As one analysis put it, for startups and enterprise buyers this is a procurement problem, not just a Microsoft problem. When pricing structures become as variable as cloud infrastructure — “powerful, variable and capable of surprising you at the end of the month” — regulators are likely to see opaque communication as a consumer protection risk, especially when customers are defaulted into higher‑cost plans.
What This Means for Buyers and Vendors
Microsoft’s situation is a warning for every tech vendor selling enterprise software and AI: price changes are no longer judged only by internal cost models, but by how clearly they are communicated and how much room customers have to say no. The AGCM’s view that Microsoft’s communication “may also constitute an aggressive practice” because it restricted consumers’ freedom of choice captures the new regulatory mood. Silence, fragmentation and default opt‑ins are starting to look like legal liabilities, not clever go‑to‑market tactics. On the buyer side, the lesson is equally blunt. Enterprises can no longer treat AI features as throw‑ins on a harmless subscription. They must ask what a normal task costs, what an expensive one looks like and how to stop runaway usage from turning into budget shocks. Vendors will keep pushing metered models because the economics demand it. The question now is whether they will accept that software pricing transparency is not optional, and that the next investigation may hit those who treat communication as an afterthought.
